It's a Tuesday morning, and you're getting your toddler ready for daycare. But as you brush her hair into a ponytail, you notice that her cheeks are flushed, her nose is runny and her breathing is phlegmy. Instead of daycare, you head to your pediatrician, who offers a solution: a prescription for antibiotics.

This scenario has played out many times in clinics around the world. The problem is that most common childhood infections are caused by viruses, which antibiotics don't treat. In the long term, unnecessary antibiotic prescriptions can foster resistance in bacteria, thereby reducing the drugs' potency and fueling the rise of dangerous superbugs.

That's why, in recent years, there's been a huge push in many countries to raise awareness of antibiotic misuse among doctors and to reduce the overprescription of these drugs. In Japan, the government noticed that the country's pediatricians were prescribing too many antibiotics, especially to young children, and offered them a financial incentive to break the habit.

It's been working remarkably well. Among young kids directly affected by the incentive, it slashed antibiotic prescriptions by 20%; it also contributed to a spillover effect in older kids and teens, as under-20s have seen a 50% reduction in prescriptions in recent years. I went to Japan to examine why the program has been so effective and to learn whether similar approaches could work in the U.S.

Understanding why a fairly modest incentive of 800 yen (about $5) a visit is so effective requires unpacking why Japanese doctors overused antibiotics in the first place. If doctors face similar pressures stateside, maybe the same strategy could work here too.

A culture of overprescription

Japan started out with a big problem of antibiotic misuse, especially among prescriptions for kids. From 2013 to 2016, over 30% of children with respiratory tract infections were given antibiotics, despite most of those infections being viral. (The U.S. started at a slightly better baseline regarding antibiotics given for respiratory conditions, with about 23% of outpatient prescriptions being unnecessary.)

For decades, it was common in Japan to routinely prescribe antibiotics for sniffles and tummy bugs, Dr. Takemi Murai, deputy head of the Infectious Diseases Division at Nagano Children's Hospital in Azumino, told me. That's in part because medical education there didn't emphasize the risk of "superbugs," meaning bacteria that can withstand many antibiotics.

"When I trained, antimicrobial stewardship wasn't yet a formalized part of the curriculum in the way it is now," said Murai, who finished medical school in 2006.

There also may be a generational divide, Dr. Tatsuki Ikuse, a pediatric infectious disease specialist at the National Center for Child Health and Development in Tokyo, told me. Ikuse, who finished medical school in 2013, suspects that older doctors "experienced many cases of bacterial infections when there was no vaccination."

Now that vaccines for whooping cough, diphtheria and pneumococcal disease are routine, children face a lower risk from the bacteria that cause those illnesses. But those early experiences treating children with serious bacterial infections may still be shaping doctors' prescribing patterns, even decades later, Ikuse thinks.

Surveys conducted in Japan suggest that doctors under 50 are more likely than doctors over 50 to see antimicrobial resistance as an urgent issue. Clinic doctors — those likeliest to see kids as outpatients for acute illnesses — also skew older than doctors in hospitals; in 2024, clinic doctors averaged about 60 years old, while those in academic and nonacademic hospitals averaged 40 and 48 years old, respectively.

Historically, Japanese doctors also had a strong bias toward prescribing antibiotics for "feverish" diseases, noted Dr. Norio Ohmagari, director of disease control and prevention at Japan's National Center for Global Health and Medicine.

Murai agreed with that sentiment. "If doctors saw a fever and a high CRP level [a sign of inflammation], they prescribed antibiotics," Murai said. "Doctors were following such a practice for a long time, so they didn't question it."

Concerns over secondary infections were also a big driver of unnecessary antibiotic use, said Dr. Masahiko Sakamoto, a hospital pediatrician in Saku, a highland city about an hour northwest of Tokyo by bullet train. Even if doctors suspected a viral infection, like the common cold, they worried that the illness could open the door for secondary bacterial infections, such as bacterial pneumonia. However, recent studies suggest this scenario is uncommon and shouldn't prompt doctors to use antibiotics preventatively.

Is antibiotic misuse still a problem in the U.S.?

Young children in the U.S. were once the most likely group to receive unneeded antibiotics, but now, those unnecessary prescriptions have "dropped dramatically," Dr. Sarah Kabbani, director of the U.S. Centers for Disease Control and Prevention's (CDC) Office of Antibiotic Stewardship, told me in an email.

Between 2011 and 2016, children's antibiotic prescriptions decreased 13% overall. "Overall declines in antibiotic use over the past decade have been driven in large part by reductions in prescribing for children," Kabbani noted.

Japan has also decreased its overall antibiotic use in younger populations, halving its outpatient prescription rates for patients under 20 between 2011 and 2022, Dr. Yusuke Okubo, division chief of clinical epidemiology and health services research at the National Center for Child Health and Development in Tokyo, told me.

The decrease in antibiotics overuse syncs with parents' lived experiences in both countries.

"I was taking antibiotics quite often" as a kid, Tatsuya Kanno, a software engineer and father of two in Tokyo, told me. "Nowadays, we don't get really prescribed those antibiotics easily."

"It has definitely been different from my childhood," said Gabby Brown, an apparel consultant and mother of two in Littleton, Colorado. "There was always a bottle of pink stuff in the fridge, you know? And now they're very hesitant to give it out."

Brown's pediatrician actually has an explicit policy on their website stating that they don't overprescribe antibiotics and instead follow evidence-based guidelines to determine when the drugs are truly necessary. I've found a number of other U.S. pediatric clinics that do the same.

One reason U.S. pediatricians have been able to improve several key metrics of antibiotic use is that the issue has been on their radar for a while.

"Antibiotic stewardship was a core tenet of my residency training 20 years ago," Dr. Morgan Leafe, a U.S.-based pediatrician who worked in inpatient and outpatient settings for 11 years after residency, told me in a direct message. "So I would say it's not new in pediatrics."

But there remains room for improvement.

Prescription rates vary widely among states and among different types of outpatient facilities, such as community practices versus those affiliated with academic medical centers. Rural communities often use antibiotics at higher rates and log more inappropriate use, compared with urban areas. These regional variations are "not explained by clinical factors," said Julia Szymczak, a medical sociologist and associate professor in the Division of Epidemiology at the University of Utah School of Medicine.

Nowadays in the U.S., urgent care clinics might be a hotspot for antibiotic misuse. These and other walk-in clinics serve an increasing number of pediatric patients, with an estimated 26% of children visiting one in the past 12 months. Urgent cares may prescribe antibiotics to children at a higher rate than other ambulatory settings do; in one study, 42% of the children seen at urgent cares were prescribed antibiotics at their visit, while the national average across all pediatric ambulatory visits is 21%.

It may be that urgent cares are more likely to prescribe antibiotics for conditions that they're explicitly not recommended for. One national analysis found that urgent cares prescribed antibiotics for 20% of allergy cases, 42% of viral upper respiratory infections and 52% of uncomplicated ear infections, while doctor's offices wrote far fewer of these inappropriate prescriptions. Other studies highlight similar patterns.

Among children, inappropriate prescriptions are less likely at urgent cares that specialize in pediatrics, highlighting doctors at general clinics as a potential population to target. Additionally, research has found that visits with nurses or physician assistants are more likely to end in an antibiotic prescription than those with doctors. (Only licensed doctors can write prescriptions in Japan, so this factor is more unique to the U.S.)

Different systems, similar pressures

Japan's nationalized healthcare provides a staggering degree of choice, because insurance coverage doesn't tie you to specific clinics or require you to get referrals to access certain care. Young children's care is virtually free thanks to national and local subsidies. So on paper, parents can easily shop around for a pediatrician that best suits their family's needs, as well as pop into any healthcare facility with flexibility should the need arise.

While this system is convenient, it has its downsides, some parents told me.

"In the U.S. and in Switzerland, they try to get to know you and know your child," said Reid, a fundraiser and father of two based in Okinawa, who recently moved to Japan with his sons and husband. "They don't at all in Japan."

In contrast, about 9 in 10 children in the U.S. have a primary care provider, although that percentage has fallen in recent years as the use of urgent care has increased. Primary care providers can be a go-to for sick visits as well as an ongoing monitor of kids' health via annual check-ups. (Annual check-ups in Japan are handled by local municipalities, not by children's regular doctors, Reid noted.)

In some ways, pediatric clinics in Japan can be considered more analogous to urgent care in the U.S., in that they're intended to address acute cases of illness as they emerge. For more serious illnesses, caregivers bring their children to hospitals, parents in Japan told me.

Diagnostic uncertainty

For a typical respiratory or tummy bug, doctors in both the U.S. and Japan make diagnoses based largely on a patient's symptoms and timeline of illness. In the U.S., most minor infections are "generally assumed to be viral unless certain criteria are present," Leafe said. These criteria are spelled out in manuals used by pediatricians.

The criteria enable doctors to "feel reassured" that they've correctly classified an infection as viral or bacterial, even in the absence of a highly sensitive diagnostic test, said Dr. Ilan Shapiro, a community pediatrician at AltaMed Health Services, a nonprofit health network in Southern California.

Rapid tests are available for only a handful of bacteria, including Streptococcus pyogenes (also called group A strep), a common culprit behind strep throat and scarlet fever. These tests are "definitely underutilized" by Japanese clinics, Ohmagari noted, while U.S. clinicians are encouraged to use them more widely.

"For strep throat, we never treat [give antibiotics], now, without a positive test," said Dr. Jennifer Shu, a pediatrician with Children's Medical Group, a group of pediatric offices in the Atlanta metropolitan area. (That said, electronic medical records suggest that not all doctors' offices use these tests before prescribing antibiotics, despite these recommendations.)

In both countries, doctors can use rapid tests to diagnose common viral infections, such as influenza and RSV. But those tests don't always catch every case. And even if they come back positive, they can't rule out the possibility of simultaneous infections with viruses and bacteria, Shu noted.

Shu's practice also uses a 20-minute test that looks for multiple viruses and bacteria in the nose and throat, including the bacteria behind whooping cough and Mycoplasma, which causes "walking pneumonia." Not all clinics use this type of test, though, because they require a special certification and equipment to run on-site, Shu said.

Such panel tests can be helpful but also tricky to interpret. Bacteria can sometimes be detected in a patient's nose without being the cause of their symptoms. "That's where clinical judgment comes in," Shu said, emphasizing that a doctor must combine test results with the timeline of a patient's symptoms.

Time crunches

The lack of quick, surefire diagnostic tests for bacterial infections can leave doctors with a degree of uncertainty, which can sometimes prompt unnecessary prescriptions, Szymczak said.

"Our number one motivation is to not harm somebody; that's the main thing," said Dr. Shruti Gohil, an infectious-disease specialist at UCI Health who designs interventions to improve doctors' antibiotics use. Worries about potential missed diagnoses can prompt doctors to reach for an antibiotic now "in exchange for a potential future, untoward consequence," such as a greater chance of resistant infections, she said.

In both countries, doctors must make diagnoses and treatment plans in a limited amount of time. "Pediatric outpatients in Japan are very busy and can only secure a few minutes for each patient," Ikuse told me.

Such time crunches are common in the U.S., too.

"I had a pediatrician who said they had — I can't remember the figure — but it was like 800 seconds for a sick visit," said Szymczak, who studies the factors that drive clinicians to misuse antibiotics. "They broke it down into seconds."

Given that antibiotics are generally very safe drugs, pediatricians can be tempted to prescribe them "just in case" in these situations, even if the need isn't certain, a 2019 review found. In these scenarios, patients end up taking courses of antibiotics they likely don't need.

Doctor shopping and bad reviews

One factor driving antibiotic misuse that might be unique to Japan is the competition between clinics and the customer service culture it motivates.

Japan's mandatory national health insurance makes it easy to "doctor shop." Depending on where a family is based, there may be a glut of pediatric specialists to choose from — as in metropolitan centers like Tokyo — or very few. For instance, in Okinawa, general family doctors are easier to come by than pediatricians, said Sandra Miller*, a researcher and mother of one based in the prefecture.

But across all settings, families can still flexibly choose which clinic they go to. Meanwhile, pediatricians in Japan, who make less than their peers who care for adults, can worry about losing clients to nearby practices.

Japanese clinicians used to worry that, after being denied antibiotics for their sick kids, dissatisfied parents would leave poor reviews online — particularly in Google Reviews — and thus drive away business, Okubo said. "They feared such clinic reviews."

Parents frequently used to demand antibiotics from their children's pediatricians, Okubo added. "Ten years ago, it was a common situation," he told me.

Among the clientele of Shu's Atlanta clinic, she's "seeing a trend where people aren't asking for antibiotics as much." But still, parents sometimes request the drugs when she's determined they're not needed. These requests often stem from the parents' past experiences.

"They'll say, 'Well, last time they took an antibiotic they got better faster, the next day,'" Shu said. Because mild viral infections often go away on their own in a few days, that timing could have just been a coincidence, but it leaves an impression on the caregiver nonetheless.

A parent in Japan told me about the other side of this conversation.

Kanno once brought his daughter into a clinic after she developed a bad cough after catching the flu. Their doctor prescribed antibiotics, and the cough resolved within days. Later, Kanno's son had similar symptoms, but the doctor didn't recommend an antibiotic in his case. When his son's coughing and wheezing persisted, "I asked him to prescribe that antibiotic, the same one that my daughter took," Kanno said.

The pediatrician said it likely wouldn't help but didn't explain why. The doctor said he would provide the antibiotic if the family wanted it. He ultimately prescribed it, and it didn't help. "He got my trust after that," Kanno said.

Szymczak's research has identified this "it-helped-last-time" bias as a common driver of patients' antibiotic requests that can be difficult for doctors to negotiate.

Parents' understanding of antibiotic resistance

In Japan, parents' awareness of antibiotic resistance has grown in recent years, Japanese doctors told me, but research suggests many still don't know much about it.

"I don't think parents generally understand that antibiotics don't work for colds, much less [understand] antibiotic resistance," said Dr. Yusuke Shibata of the Shibata Pediatric Clinic in Tokyo. "Explaining antibiotic resistance to parents is difficult," though he still tries to do so, he told me in an email after I visited his clinic.

Studies of U.S. parents suggest that many don't understand resistance or proper antibiotic use and that a misunderstanding of the purpose of antibiotics can drive requests for the drugs. In my interviews, I found that parents in both countries had some grasp of the concept of antibiotic resistance, but the depth of that understanding varied.

Tomomi Sato*, a New York City-based teacher and mother of two who lived in Japan as a child, said she'd heard that taking antibiotics too often could make it so your body didn't respond to them well in the future. She likened it to an urban legend about roaches that she heard from her mother and others from her hometown of Tokyo.

"You know how they talk about, in Japan, roaches are getting bigger and bigger?" she said. "They just build immunity … they don't respond to the old type of roach spray, so you have to get a new version."

Some parents are cautious about overusing medications in general but don't worry about antibiotic resistance, specifically.

"I want them to recover by trusting their own immunity, rather than relying on medication," said Risa, an HR representative and mother of two in Saku, Japan. But that said, "if it's something that's difficult to heal naturally, I think professional judgment and a prescription are necessary," she added.

Risa told me she thinks antibiotics may be medically appropriate when an illness is particularly severe or prolonged. She doesn't weigh whether the cause is likely viral or bacterial, even though antibiotics don't treat viral illnesses.

"I think that's probably the general mindset," said Fumie Kuchiba, a mother of one in Saitama, Japan, who previously worked as a registered pharmaceutical salesperson.

Other parents have a clear sense of the threat resistance poses.

"Resistance is a threat because if something is powerful enough to beat our current modern medicine, then it can wreak real havoc," said Malcolm, an editor and father of one in Atlanta. On the individual scale, he knows that resistance can emerge within an individual, potentially raising their personal risk of resistant infections. But on a global scale, he thinks of resistance as a threat as formidable as that posed by viral pandemics, such as COVID-19.

Difficult conversations

When parents seek an antibiotic that isn't medically needed, that can prompt difficult conversations with their children's doctors.

A Philadelphia-based doctor summed up the dynamic in an interview with researchers: "If patients feel like they need an antibiotic, it's very, very hard to talk them down from that. So, antibiotics are definitely overprescribed. And, you know, it's easier to say yes rather than taking time and, like, so much time to say no." This doctor and several other participants in that study felt they had "little autonomy to stand up to patient demand."

Some studies find that, when a pediatrician withholds antibiotics and a parent questions them, doctors can perceive that as negative pushback and give the drug against their initial judgment. But if a pediatrician denies antibiotics while also providing guidance for symptom relief, that conflict can be avoided and prescription becomes less likely.

Shapiro, the pediatrician in California, told me he handles such requests by walking parents through his diagnostic process and explaining the downsides of using unnecessary prescriptions. He also offers options for seeking further care — via phone, telehealth or an in-person visit — should the infection not improve as expected.

These conversations take time, and they can stir up conflict, which Shapiro said he's grown more comfortable navigating over his career. Early on, "you don't want conflict; you feel that conflict is bad, and you want everybody to be happy," he said. But ultimately, "the objective is for them to be healthy."

Multiple parents in Japan told me their doctors don't usually talk through their reasoning for checking for a given symptom or prescribing X drug over Y. "If we don't ask, he won't explain," Kanno said of their current pediatrician; but if prompted, their doctor provides good explanations, he said.

Similarly, Japanese parents told me their doctors don't often say whether or why they suspect an infection is viral or bacterial — a clarification that might help caregivers feel more comfortable forgoing an antibiotic.

Notably, Japan's antibiotic incentive program, which allows pediatricians to earn "tips" for withholding unnecessary antibiotics, provides the payment only if the doctors also explain responsible antibiotic use to children's caregivers. That requirement may be necessary to prompt that conversation.

Some doctors in Japan who claim the incentive provide verbal explanations about resistance or give parents a slip of paper that explains the basics and includes links to further information. Dr. Atsushi Miyahara of the Karugamo Clinic in Tokyo told me that, over time, his clients have become more knowledgeable about antibiotics, and insurance data related to the incentive also hints that this learning takes place. But it's unclear if these interactions can always move the needle.

Fielding parents' concerns

Shu, the Atlanta-based pediatrician, finds that some parents ask for antibiotics because they think the drugs will speed their child's recovery, without considering whether a bacterium is the culprit. Caregivers in both the U.S. and Japan also told me that they know parents who seek antibiotics as a quick remedy for illnesses.

Parents I spoke to who understood that antibiotics only treat bacterial infections added that it can be frustrating that there aren't many medicines available for viral ailments. Waiting for an infection to clear up can be stressful because you don't want your child to suffer, parents said, and the sickness also disrupts the family's normal routine of work, school or daycare.

"When your kid gets sick, things just kind of fall apart, and you need your life to get back to normal," said Katie, a journalist and mother of one in Jacksonville, Florida, who added that many parents may lack support that would help them take time off to care for their kids.

In both countries, pediatricians feel pressure to leave caregivers satisfied. While clinics in the U.S. aren't necessarily competing for patients in the same way Japanese clinics are, there is still a degree of customer service at play, Szymczak said. "That clinical encounter is very transactional, particularly in the United States," she said.

Doctors want to provide something of value — a diagnosis, medicine, reassurance — in exchange for the caregiver's time and money. Sometimes, pediatricians assume that what the caregiver wants is antibiotics, when really, they want a treatment plan to help their child, Szymczak said.

Could incentives motivate change?

While the U.S. has made progress in curbing antibiotic misuse, there's room for improvement. Could we close those gaps by using a similar approach to Japan's 800-yen incentive?

Similar dynamics are at play in both countries, with parents and pediatricians in both places dealing with similar issues. But while these pressures are similar in the clinic, a key difference sets America's situation apart: Our complex healthcare system, funded through a patchwork of insurers, can be difficult for patients and doctors to fully trust.

In the final installment, I'll talk with U.S. doctors to see what they think about Japan's approach and explain solutions that would suit our own fragmented healthcare system.

Editor's note: The names of some parents quoted in this story have been abbreviated or changed to protect their privacy. Altered names are marked with asterisks (*). This article was updated on Aug. 4, 2026, to note that the panel test used at Shu's practice takes about 20 minutes to run, not 45 minutes.

About a decade ago, the Japanese government spotted a worrying pattern: Pediatricians were doling out a ton of antibiotics, well beyond what should be needed to treat the bacterial infections coming through their doors.

Antibiotics treat bacterial infections, not those caused by viruses, fungi or parasites. Yet doctors were often sending young patients home with antibiotics for illnesses unlikely to be bacterial. Treating nonbacterial infections with antibiotics can fuel antimicrobial resistance (AMR) and the rise of superbugs by unnecessarily exposing bacteria to the drugs, pressuring those bacteria to evolve strategies to survive. Resistant bacteria can then spread their adaptations to others, thereby compounding the problem.

In the long run, resistance could make common infections impossible to treat with existing drugs, raising the risk of serious illness and death across the population.

When pediatricians in Japan did treat bacterial infections, they were overusing the antibiotics that are likely to fuel resistance — "broad-spectrum" drugs that target many bacteria at once. When compared against 35 other high-income countries in 2015, Japan ranked dead last in the appropriateness of antibiotic choices for kids under 5.

One way to slow the development of AMR is to get antibiotic prescriptions under control. So Japan focused on one of its biggest sources of problematic antibiotic use: pediatric outpatient clinics.

"The clinics are a particular problem," said Dr. Yusuke Okubo, chief of clinical epidemiology and health services research at the National Center for Child Health and Development, a research center and hospital in Tokyo. Looking at Japan's overall antibiotic use, "90% of prescriptions are [from] outpatient clinics, not hospitals," Okubo told me. (Outpatient practices account for a similar proportion of antibiotic prescriptions in the U.S.)

A large proportion of problematic prescriptions were being written for children under 3 years old, especially those with upper respiratory tract infections, which nine times out of 10 are caused by viruses, Okubo said. The stomach bug gastroenteritis was another biggie for overprescription, despite most often being caused by a virus.

Government officials hatched an idea: What if each time a pediatrician chose not to prescribe an antibiotic in these cases, they earned a small financial reward — a tip for making a better choice? Each tip would be small, amounting to about $5 per claim at today's exchange rate, but could translate to thousands of extra dollars of annual income for individual clinics. That's no small matter for Japanese pediatricians, who Okubo estimates earn roughly $90,000 to $100,000 a year.

This incentive program, rolled out in 2018, has proved a success, so much so that it's since been expanded to cover more patients, more clinics and a wider variety of diseases. But what made the program work? I went to Japan to find out what systemic and cultural factors contributed to the program's success. Ultimately, I wanted to see whether other countries, like the U.S., could learn from this nationwide experiment.

The quietly rising threat of resistance

With bacteria, as with animals, it's survival of the fittest. When exposed to antibiotics, a percentage of bacteria die, while others survive. That surviving population has traits that help them withstand the drug's effects, which are encoded in "resistance genes." Bacteria can transfer those genes to the next generation by multiplying, as well as physically pass those genes to nearby bacteria. Plus, they can pick up new resistance genes through random DNA mutations.

All antibiotics come with the risk of pressuring bacteria to evolve resistance — it's an inherent feature of the drugs. However, broad-spectrum drugs carry the greatest risk, because they place pressure on a wider variety of bacteria than narrow-spectrum drugs do. The 2015 cross-country comparison found that Japanese doctors prescribed antibiotics with the lowest risk of resistance only 35% of the time, meaning most antibiotic prescriptions were for broader-spectrum drugs.

It's sometimes necessary to use broad-spectrum drugs — such as when an infection is resistant to narrow-spectrum options — but using broad-spectrum antibiotics when they're not needed hastens the development of AMR. So does using antibiotics for nonbacterial infections. In both scenarios, you're introducing evolutionary pressure that could have otherwise been avoided.

Globally, resistance is rising among common disease-causing bacteria; it's far outpacing the development of alternatives to antibiotics. In 2021, resistant bacterial infections directly caused 1.14 million deaths worldwide and contributed to another 3.57 million deaths. Those numbers could climb dramatically by 2050 if swift action isn't taken now.

Common pathogens already show high rates of resistance in Japan, and resistant germs contribute to thousands of deaths each year, with most occurring among older adults. People over 65 make up about 30% of Japan's population. As that percentage grows in the coming years, AMR-related illness may also increase, health officials worry.

Japan's youngest residents are also at risk, though, as resistant bacteria can sometimes pass from mothers to newborns at birth and can cause serious complications, like sepsis. Resistance has also been detected among respiratory bugs that frequently infect kids, such as Mycoplasma pneumoniae, said Dr. Takemi Murai, deputy head of the Infectious Diseases Division at Nagano Children's Hospital in Azumino. "There have been outbreaks of Mycoplasma that are resistant to antibiotics," he said.

Yet not long ago, the unrestrained use of antibiotics was a mainstay of Japanese medicine. (I'll dig into the myriad reasons why in later installments of this series.)

National insurance data sampled from 2005 showed that 60% of patients in Japan with nonbacterial upper respiratory tract infections were prescribed antibiotics, mostly broad-spectrum ones like third-generation cephalosporins, macrolides and quinolones. Most of those prescriptions came from clinics.

Something had to shift.

Tips for appropriate treatment

In 2016, Japan got serious about reducing its antibiotic misuse, releasing its first National Action Plan on Antimicrobial Resistance. It aligned with a global plan from the World Health Organization, which aimed to raise awareness of AMR and optimize the use of antimicrobials, including antibiotics, among its member states.

Two of Japan's big goals were to slash overall antibiotic use by 33% and broad-spectrum antibiotic use by 50% by 2020. The country came very close to hitting those ambitious targets by the deadline, and doing so was no small feat. The tipping program was just one of a slew of initiatives introduced to improve AMR awareness and antibiotic use.

Prior to the incentive's introduction, Japan's Ministry of Health, Labour and Welfare (MHLW) — whose role is somewhat analogous to the U.S. Department of Health and Human Services — rolled out educational campaigns for doctors and patients and wrote a manual for antibiotic use, with the first edition directed primarily at outpatient doctors. It emphasized that most acute respiratory tract infections and acute diarrheal diseases don't require antibiotics.

Among kids, children under 5 saw the highest antibiotic prescription rates, often receiving the drugs for respiratory infections. Data showed this overprescription problem was the worst in children under 3 being assessed for upper respiratory infections or acute gastroenteritis. The government's solution? Pay doctors extra to withhold antibiotics when faced with cases that don't warrant them.

"If the clinicians provide more appropriate medical services, we add additional payment," said Dr. Takuma Kato, a counselor at the Permanent Mission of Japan to the United Nations who previously worked on the incentive program for MHLW. In this case, they pay "a little bit more" when doctors don't give patients antibiotics for illnesses that are likely viral, he said.

A "little bit more" is accurate. Each tip is 800 yen, equivalent to about $7.20 when the program launched in April 2018 and about $5 at today's exchange rate.

Here's how it works: A caregiver brings in their sick infant or toddler for an initial visit, and the pediatrician determines the child likely has an acute upper respiratory tract infection or gastroenteritis. These illnesses are typically caused by viruses, so the doctor decides not to provide an antibiotic. The doctor explains this rationale to the caregiver and provides guidance for home care. If the appointment checks those boxes, the clinic can claim an extra 800 yen when they seek reimbursement.

Because mild viral infections typically resolve on their own in a few days, a strategy called "watchful waiting" can help clarify if a bacterium is actually at fault. So if a doctor does prescribe an antibiotic at the first visit, they'll often encourage the caregiver to bring the child back if their condition remains the same or worsens within a few days. The incentive and this "waiting" strategy go hand in hand.

There are a few technicalities. For example, the children being assessed must have no underlying conditions that might complicate their case, such as a weakened immune system. If children test positive on a formal diagnostic test for influenza or COVID-19, the incentive cannot be claimed. To qualify for the incentive, clinics must specialize in pediatrics and use a "comprehensive" payment system, meaning patients pay a standardized amount for the whole appointment rather than the itemized "fee for service" that is ubiquitous in the U.S.

Despite this fine print, the perk is pretty appealing for the clinics that claim it. "I think pediatric doctor associations are really happy," Okubo said.

Doctors say "small incentives add up"

The Japanese government generally incentivizes doctors to adjust their behaviors, rather than penalizing them for poor practices, Okubo said.

"It's a constructive message from the government: 'You changed your behavior, so we'll pay something,'" Okubo said. "This constructive approach motivates physicians, especially pediatricians, to apply their common sense to their actual practice."

This system is readily accepted by Japanese doctors, who have historically held a lot of political power, Kato noted. Just like in the U.S., their professional groups, such as the Japan Medical Association, lobby the government and typically push against policy proposals that they view as potential threats to their bottom line.

By contrast, an incentivizing approach is "very, very welcome, especially by the doctors' associations," said Dr. Norio Ohmagari, director of disease control and prevention at the National Center for Global Health and Medicine, part of the Japan Institute for Health Security (JIHS) in Tokyo. Ohmagari also leads the AMR Clinical Reference Center, which collaborates with the WHO on AMR countermeasures.

Doctors I asked confirmed that they like the incentive, namely because it boosts their profits.

"I apply for the pediatric antibiotic appropriate use support premium each time" it's applicable, said Dr. Yusuke Shibata, who runs the Shibata Pediatric Clinic in Asakusa, a historic district in Tokyo's Taito ward. "I appreciate the premium, as pediatric clinics already have low profits" compared with clinics that care for adults, Shibata told me in an email after I visited his clinic.

For first visits with kids under 6, clinics are paid a base rate of 6,040 to 7,210 yen, or about $38 to $45 at current exchange rates. An extra 800 yen (about $5) increases that fee by more than 10% — "a huge amount," Okubo emphasized.

Shibata estimates that his clinic sees about 30 to 40 patients with an acute respiratory infection or diarrhea each week, depending on the season. He can potentially claim the incentive for the first visit with each of these patients, assuming they don't have any conditions that would disqualify the claim.

On the high end, Shibata estimates that he might claim the incentive 180 times in a single busy month, which would total 144,000 yen, or about $900.

Dr. Atsushi Miyahara, who runs the Karugamo Clinic in the Setagaya ward in Tokyo, said he was already conservative about using antibiotics so the incentive rewards him for sticking with his status quo.

Fifteen years ago, when Miyahara opened his clinic, he noticed that other physicians prescribed a lot of antibiotics, and he questioned the practice due to the potential to fuel resistance. He provides his patients with informational flyers that explain the risks of resistance and how avoiding unnecessary prescriptions can reduce that risk. When antibiotics are needed, he predominantly uses narrow-spectrum drugs that pose a relatively low risk of resistance.

Miyahara said the local government and medical associations announced the antibiotic incentive when it was launched, and he felt its introduction has been very positive. It's increased his revenue and encouraged him to continue his stewardship practices. He estimates that for every 50 of his first visits with patients, he claims the incentive 10 to 15 times, so it applies to at least 20% of those visits.

The incentive's impact

Because some pediatric clinics qualified and others didn't — due to fee-for-service clinics being excluded — Okubo and his colleagues could directly measure whether the policy worked.

To assess the incentive's effects in its first year, the researchers looked at insurance claims from over 10,000 medical facilities from just before and after the policy's introduction. About 3,000 of the facilities qualified and claimed the incentive 316,770 times, totaling 253 million yen ($2.29 million at the time). These eligible clinics saw a 17.8% reduction in their total antibiotic use over a year without any negative effects for patients, such as higher hospitalization rates.

To see if those effects lasted, the team drew years of data from over 165,000 young children who went to either eligible or ineligible clinics. Within the first month of the policy's implementation, the former group of children saw a nearly 45% reduction in total antibiotic prescriptions, compared with the other kids. Cumulatively over the next four years, their overall antibiotic use and broad-spectrum-antibiotic use was 20% and 24% lower, respectively.

The decrease in antibiotic prescriptions did not come with an uptick in hospitalizations or healthcare costs, although there was a slight increase in the total number of doctor's visits. But that's what you'd expect as doctors track an infection over time, Okubo explained, meaning they likely employed the watchful-waiting strategy and had parents bring their kids back in if they didn't improve quickly.

Okubo's team has continued to track pediatricians' antibiotic use, and he noted that they're seeing "spillover effects" among age groups not covered by the incentive. In the under-20 age group, outpatient antibiotic prescriptions fell by 50% between 2011 and 2022. He thinks the incentive is a key driver of this trend, directly reducing prescriptions for the youngest kids while also triggering ripple effects in older groups. (This research will soon be published in a peer-reviewed journal.)

That said, there's room to improve doctors' selection of antibiotics when they are used, as the ratio of broad- to narrow-spectrum drugs is still too high. "Total antibiotic use was reduced, but its quality should be improved further," Okubo said.

The incentive's evolution

While the 800-yen tip has proved significant to clinicians, the incentive represents a very small slice of overall government healthcare spending, which totaled 468 billion yen ($3.1 billion) in 2022.

"This program is not large compared to the whole budget," Dr. Kosuke Sasaki, who works in the MHLW's health insurance bureau, told me. The program's budget has no upper limit, so if the number of claims from clinics increased, there isn't a cap on how many could be paid out. "The number of doctors using this program has increased while the number of antibiotic prescriptions has decreased since the start," Sasaki's colleague Dr. Tomonori Aoki added, noting that the government isn't concerned about how to pay that rising bill.

The program's measurable impact and low price tag may explain its growth over the years.

Every two years, Japan's Ministry of Finance hands the MHLW its slice of the government budget, and MHLW then revises the pricing for drugs, medical devices and healthcare services. The antibiotic incentive falls under this revision process and has been expanded several times.

During its first revision in 2020, the incentive was extended to children under 6. In 2022, ear, nose and throat specialists (ENTs) newly qualified for the incentive; like pediatricians, they treat many acute infections in kids and tend to overuse antibiotics, insurance data suggested. That same year, doctors could start claiming the fee for ear infections and sinus infections.

"I see a tendency for pediatric clinics to avoid prescribing antibiotics, but I do see some ENT clinics prescribing antibiotics carelessly," Shibata, the clinic owner in Asakusa, Tokyo, told me. So ENTs seem to be a logical next target.

In 2024, a separate, facility-level incentive was introduced as a complement to the 800-yen incentive. It encourages clinics to submit data to a government database that tracks antibiotic use. If first-line, narrow-spectrum antibiotics make up a certain percentage of the clinic's overall prescriptions, that clinic earns extra money, Okubo explained.

Ultimately, the 800-yen incentive helped put AMR on the radar of doctors who didn't take it as seriously as experts like Kato, who told me "AMR is kind of my life's work." Kato and researchers like Okubo see the program as a success, although they pointed to room for improvement in antibiotic selection. Ministry officials like Sasaki and Aoki said the program is easy to implement and makes a difference. Clinic doctors like Shibata and Miyahara appreciate the incentive and use it consistently. All in all, at just $5 a claim, the incentive has been remarkably effective.

In speaking with experts in Japan and the U.S., I've learned that U.S. doctors have historically faced the same pressures and showed similar lapses in antibiotic use that Japanese doctors have. However, the U.S. does not have an incentive program like Japan's. Should it launch one?

In the next installment of this series, I'll explore a central feature of this problem: the motivations behind pediatricians' antibiotic misuse. What are they, and do those motivations differ between doctors in the U.S. and Japan? Answers to those questions will help determine whether a similar incentive might have the same impact in both places.

China controls some of the largest rivers in Asia, experiences heavy rain and has vast stores of water locked in glaciers in the west. But despite these abundant resources, China has a water problem. While central and southern regions are quenched by massive rivers like the Yangtze, the northeastern megacities like Beijing and Tianjin have endured serious water scarcity as agriculture, industry and population size rapidly increased.

To address the imbalance, China has undertaken one of the biggest engineering projects in the world: rerouting the flow of one of the world's mightiest rivers and its tributary, and ferrying the water thousands of miles from the center and south of the country to the thirsty north.

Called the South-to-North Water Transfer Project (SNWTP), the vast system of canals, pipes, dams, reservoirs and pumps moves staggering amounts of water via two routes through central and eastern China.

"It binds into a single network four major river basins, six provinces, three megacities, myriad bureaucracies, and more than 700 million people," Michael Webber, a professor emeritus in the School of Geography, Earth and Atmospheric Sciences at the University of Melbourne who specializes in regional economic development in China, wrote in a 2023 journal article.

But even this is not enough to satisfy the water needs in the North: China is currently expanding these routes and is planning a third, western route for the project, which would snake through the Tibetan Plateau before feeding the water-parched north and east. The route is attractive to Chinese officials because the river-and-glacier-threaded plateau, dubbed the Water Tower of Asia, has abundant water resources and is the source of numerous major rivers.

However, experts say the routes could worsen conditions in parts farther south that provide the water. What's more, the Tibetan portion of the project has stoked fears and suspicion in neighbors, as unofficial plans swirl suggesting major transboundary rivers could be diverted away from other countries and further into China.

The huge plan is "unbelievable," Mark Wang, a professor of human geography at the University of Melbourne, told Live Science. "The proposals [have] underestimated the negative impacts — environmental, earthquakes, evaporation and economic cost," as well as the international impacts, Wang added.

Water imbalance

Construction of the SNWTP began in 2002. The eastern route has operated since 2013 and transports water from the mighty Yangtze from Yangzhou, near Shanghai, to the megacity Tianjin, which has a population of 15 million people. The central route, which goes from Danjiangkou to Beijing and Tianjin, began flowing the following year, carrying water from the Han River, or Hanjiang. Together, they have moved 21 cubic miles (88 cubic kilometers) of water more than 1,800 miles (2,900 kilometers) since they began operating, according to the Chinese government — more than double the maximum amount of water held in Lake Mead.

The motivation for this project is obvious: China must support nearly 20% of the world's population with just 6% of its freshwater reserves. And this water is not distributed evenly across the country. Northern China has nearly half of the country's population and more than half of its agriculture but just one-fifth of the country's freshwater reserves.

In some places, the imbalance is even more dramatic. Tibet, in the west, has 4.8 million cubic feet (136,800 cubic meters) of water per person, while Tianjin, in the Northeast, has just 3,990 cubic feet (113 cubic m) per person. The United Nations considers less than 1,000 cubic m (35,000 cubic feet) of fresh water available per person water scarcity and less than 500 cubic meters (18,000 cubic feet) per person as "absolute water scarcity." The average in the U.S. is 314,300 cubic feet (8,900 cubic m).

Water transfer

China has a long history of rerouting water to serve its needs. Parts of the eastern route, for instance, use the Grand Canal, which was first built in the fifth century B.C.

And the broad concept of transferring water from the south to the north was first laid out by Mao Zedong, China's first communist leader, in 1952.

"Control of water has been a key theme in Chinese history and mythology," Tom Harper, a lecturer at the University of East London specializing in Chinese foreign policy and international relations, told Live Science.

Currently, the eastern route, which stretches 715 miles (1,150 km) from the Yangtze River to eventually reach Tianjin, transfers around 4 cubic miles (15 cubic km) of water per year.

This water needs to travel 213 feet (65 m) uphill, meaning pumping stations have to raise the water along the route.

The central route, meanwhile, relies on gravity to channel around 3 cubic miles (13 cubic km) of water 790 miles (1,270 km) from Danjiangkou Reservoir, on the Han River in Hubei province, to eventually reach Tianjin.

The Danjiangkou Reservoir already existed before the commencement of the project, but it was significantly expanded to raise the water level — and required 350,000 people to be relocated as a result.

The routes are now key to providing water to Beijing and Tianjin. For example, around 70% of Beijing's water comes from the route, experts told Live Science.

Fixing the fixes

The Chinese government and many Chinese scientists have boasted that the project has increased the water supply to people in northern China, raised the groundwater table in northern regions, and increased economic activity and agriculture in the region.

A major part of the project also focused on improving water quality: Freshwater supplies in China are heavily polluted, which reduces the amount of drinkable water.

"These were reservoirs, these were rivers, these were groundwater resources that had severe pollution impacts," Darrin Magee, a dam expert at Western Washington University, told Live Science. "So in the south where we had abundant water resources, the quality tended to be very bad in some areas."

To rectify this, the SNWTP installed water-cleaning stations along the route and the government implemented measures to reduce water pollution, including shutting down or relocating hundreds of manufacturing businesses that dumped their waste into waterways along the routes.

Along the eastern route, "pollution was terrible," Wang noted. "The central government put in a lot of effort… and they have control of the pollution again in a short period of time," he noted, adding that the water quality has significantly improved along the routes.

However, experts have said that the massive project has damaged ecosystems, displaced hundreds of thousands of people and fueled saltwater intrusion into underground reserves in water-providing areas, as the Yangtze River's freshwater flow is reduced.

"When the South-North Water Transfer was being proposed, many, many Chinese scientists were also against it because you just create lots and lots of new problems, whatever you fix," Emily Yeh, a professor of geography at the University of Colorado Boulder who specializes in Tibet, told Live Science. "Why not conserve water instead?"

Other experts agreed. "The water problem in China, it's not just physical water scarcity, it's the structural problem," Wang said. The biggest driver of water consumption is farming irrigation, which is water-intensive and leads to significant water loss. Conserving that water would go a long way to alleviating water scarcity, Wang added.

If China's "first priority is conservation, you won't need a large-scale, world's largest mega project," Wang said. Chinese authorities seem to be realizing this. They're also attempting large-scale water conservation policies, Wang said.

But the country is forging ahead with engineering projects.

The SNWTP's sources, particularly along the Han River, have less water per capita than the world's average, so siphoning off large amounts of the river's water could severely strain local areas. This has become such a problem that Chinese authorities are now undertaking even more engineering projects to alleviate the problem caused by the SNWTP. For instance, a new, smaller diversion is being added near the Danjiangkou reservoir, because people downstream are not getting sufficient water, Wang said.

The central route is further being extended to improve Danjiangkou's water levels. The central route relies on gravity, so a high water level is required for the route and the reservoir's dam to function. If water levels dip too low, as happened in 2011 during droughts, the dam stops working. So, engineers are connecting the Danjiangkou reservoir to the Three Gorges Dam, the world's largest dam, on the Yangtze, to increase the water supply.

The project is "a fix to fix the fix to fix the fix," environmental researcher Stevan Harrell wrote in the Made in China Journal — emphasizing the ever-growing number of engineering projects to deal with poor water governance.

Western route

Even though the two existing routes provide vast amounts of water to the North, and other water-conservation measures are being rolled out, Chinese authorities say the supply to the North is still not enough to meet demand, and that more water diversion is necessary.

The SNWTP's planned western route through the Tibetan Plateau won't start operations until at least 2050, but since its earliest planning stages, it has been mired in controversy.

The western route's official path would take 4 cubic miles (17 cubic km) of water per year from the Tongtian, Yalong and Dadu rivers — all tributaries that eventually feed into the Yangtze. The route would cross about 190 miles (300 km) of the Tibetan Plateau to the Yellow River, which then flows to the water-scarce north and east. The government has commissioned more feasibility studies for this route as part of the country's newly announced five-year plan.

The construction of such a massive project is fraught with difficulty. The Tibetan Plateau rises 10,000 to 15,000 feet (3,000 to 4,500 m) above sea level. These high altitudes present several problems, such as challenges in construction and maintenance, and difficulties in preventing water from freezing.

To transfer water across the plateau, the route would require pumping stations, along with a series of 300- to 1,000-foot-tall (100 to 300 m) dams, which would rank among the tallest dams in the world. The project would also require cutting 200 miles (300 km) of tunnels through the mountains, including the Bayan Har Mountains, which separate the upper Yangtze and Yellow River drainage basins.

The Tibetan Plateau is very seismically active, meaning earthquakes could lead to collapses or landslides, potentially damaging the vast infrastructure.

In 2018, Chinese scientists attempted to overcome the difficult terrain by transferring water through the air. But the "Sky River," or Tianhe project, ultimately proved unfeasible and has been effectively cancelled.

But China has not been deterred by these obstacles. The ongoing construction of the Dianzhong Water Diversion Project, in Yunnan province in southwest China, provides a test case for transferring water across mountainous, earthquake-prone landscapes. The damless design relies on a 380-mile-long (610 km) series of 58 tunnels, as well as pumps, to carry water through the mountainous and seismically active terrain. It is the "the world's longest water tunnel that could fit two high speed trains" and is "seen as a pilot demonstration project that will inspire and inform the design for the Western Routes," according to the Hong Kong-based think tank CWR. The first stage of the Dianzhong project is due for completion later this year, with a second phase recently announced.

Red flag

While the official western route has yet to break ground, alternative proposals have caused serious concern in neighboring countries. One unofficial plan developed in the 1990s, which politicians and academics subsequently considered completely unworkable, suggested funneling 48 cubic miles (200 cubic km) of water (four times the flow of the Yellow River) away from the Brahmaputra River, which flows into India and Bangladesh, and diverting the water toward the Yellow River basin, which serves the North and East of China.

A second controversial proposal, known as the Red Flag River project, suggests diverting 14 cubic miles (60 cubic km) of water 3,700 miles (6,000 km) from the Lancang, Nu and Yarlung Tsangpo rivers (called the Mekong, Salween and Brahmaputra rivers when they leave China) across the Tibetan Plateau to the northern Xinjiang region, which has high water stress and is facing desertification. Officials suggested the plan could turn Xinjiang into China's "California." The diverted water could also potentially be rerouted toward Beijing, experts told Live Science.

Some experts have called these "semi-official" proposals, as they have no official government backing but have been discussed openly by researchers with heavy government backing and support in China, which the government wouldn't allow if these people were truly speaking out of turn, Wang said.

Despite no official government backing, the plans have worried neighboring countries, particularly India, due to fears that any water diversions on international rivers could significantly reduce vital downstream supplies. Massive water engineering projects are already being built in Tibet, such as the upcoming Motuo megadam on the Yarlung Tsangpo — close to the proposed starting point for the Red Flag River project.

"There's a lot of concern that there might be water diversion, because China has been talking about this for many, many years," Tenzin Norgay, a researcher at the nonprofit organization International Campaign for Tibet, told Live Science. "If it happens, it's going to be a huge risk to downstream countries."

Researchers have calculated that the Red Flag project would likely siphon around 20% of the rivers' upstream flows, though the overall water loss for the whole of each river is less than that. Regardless of the actual amount of diversion, the lack of international consultations and transparency is exacerbating suspicions, experts told Live Science.

A changing climate

While China is forging ahead with attempts to redistribute China's water and reduce water scarcity via the SNWTP, climate change could scramble those plans. An increased risk of droughts over the coming decades could compromise the SNWTP's ability to move water.

The Tibetan Plateau, which is a source of water for almost 2 billion people, is particularly at risk. The world's "Third Pole" is rapidly warming, leading its numerous glaciers to melt. This could lead to sudden flooding, followed decades later by reduced river flow as the glaciers disappear. Desertification, meanwhile, is also impacting the Tibetan Plateau.

"Short to medium term, [we'll] see an increase in runoff as melting begins earlier each year," Magee said. "But over the long term, decreased snowpack means less secure water resources for most of China."

Taming nature

China's water management approaches include vast dam-building projects, cloud seeding, monumental tree-planting schemes and water-transfer projects, all of which rely on technological and engineering solutions to control the natural world, experts told Live Science.

When it comes to megaprojects, "If there's any country that will do it, it's China," Magee said. "The People's Republic of Engineers."

Other experts agreed. "It is part of this idea that you can engineer your way out of these problems, even if it's with trees rather than with machines," Yeh said. "A large-scale engineering approach to the natural world has been a characteristic of Chinese approaches to perceived or real environmental problems."

This is particularly true for the Tibetan Plateau. "The government is really imagining Tibet as a source of ecological service for the rest of the country," Yeh said.

Other experts agreed that Chinese authorities typically take an engineering approach to attempt to manage nature.

But Wang noted that China faces big pressures, which is why its government is looking for big solutions.

"If you understand China's energy issues, you understand why China is doing this. If you understand the water and food security issues, you understand so many things China has done," Wang said.

In southwest China, Chinese soldiers load and fire rockets toward the sky as aircraft and drones circle overhead, dropping their toxic cargo into the air. The weapons are not trained on an enemy, and the planes aren't dropping bombs. Instead, they are targeting the clouds hovering in the sky.

This footage is just a tiny snapshot of the massive, countrywide effort to seed the clouds with rain at an unprecedented scale.

More than 50 countries around the world use cloud seeding to modify the weather at small scales, including the United States.

But China is the world leader, employing around 50,000 people; using thousands of rocket launchers and dozens of planes; and investing the equivalent of billions of dollars in these initiatives, experts told Live Science.

In 2018, China embarked on its most ambitious cloud seeding plan. The Tianhe ("Sky River") project aimed to create a permanent airborne water "corridor" from one river basin to another on the Tibetan Plateau. The project was intended to be part of the country's much bigger South-North Water Transfer project — a massive effort to transfer water to China's highly populated and water-scarce North and East.

From the start, the Sky River faced scathing criticism from scientists who said the project was unworkable. Yet China forged ahead.

China's pursuit of such a scientifically questionable geoengineering technology shows just how far Chinese authorities are prepared to go to achieve the country's water and climate goals. It also reveals how the Chinese government views the natural world, experts have said.

"There is an impetus to control and to view the environment as a machine or an infrastructure that can be controlled," Emily Yeh, a professor of geography at the University of Colorado Boulder, told Live Science.

The science of cloud seeding

China uses cloud seeding to produce precipitation to build up snowpack, help alleviate droughts, reduce the impact of hailstorms, and create skies clear of clouds and pollution for official events — as it did most famously during the 2008 Beijing Olympics, experts told Live science.

But cloud seeding can't create rain from an empty sky. Instead, it causes existing clouds to generate precipitation more efficiently by injecting particles into a cloud. Water droplets then coalesce around these particles, and the water eventually falls as rain or snow.

"Every single drop in any cloud you have ever seen — there's a particle in it," Rob Rauber, an atmospheric scientist at the University of Illinois Urbana-Champaign, told Live Science. Cloud seeding speeds up the raindrop-forming process by introducing more particles, he explained.

There are two main ways to do this, depending on whether the temperature of the clouds is above or below 32 degrees Fahrenheit (0 degrees Celsius). Warm-cloud seeding is known as hygroscopic seeding, while cold-cloud seeding is called glaciogenic seeding.

"They have the same general goals or approach," Jeff French, head of atmospheric science at the University of Wyoming, told Live Science. "And that is to introduce something into the cloud that would increase the efficiency in which cloud droplets or ice crystals are able to grow to precipitation-size particles and fall out of cloud either as rain or as snow."

A salt such as sodium chloride is typically used for hygroscopic seeding to attract water droplets, French said. Glaciogenic seeding, by contrast, takes advantage of cold clouds' supercooled water — water that is liquid even when temperatures are below freezing — by introducing silver iodide. This substance has a similar structure to ice and collides with supercooled water in a cloud, causing it to freeze and eventually fall from the cloud, French said.

Scientists insert seeding particles into clouds in a number of ways, including by dropping them from planes and drones, firing shells or rockets into the clouds from the ground, or burning materials in chambers that release the combustion byproducts into the air.

Scientists are still learning exactly when and how to use these techniques most effectively. Rauber and French both noted that it's difficult to quantify cloud seeding's impact.

"It becomes kind of a quagmire when you ask that question: 'Does cloud seeding work?'," French said. For his research, he used airplanes, ground sensors and radar across the U.S. Mountain West to measure the effects of seeding from aircraft on the initiation and growth of crystals and eventual snowfall. "From a physical standpoint, I can say very confidently that cloud seeding works," French said.

However, that experiment was in cold, mountainous conditions, and the impact of cloud seeding depends on many conditions, so tracking whether cloud seeding increases precipitation over a longer period and in different conditions can be challenging.

"Not all clouds are created equal," French said. Some are colder or warmer, and the seeding material won't work as well; in others, the size or distribution of droplets makes precipitation less likely. Even within a small region, the snowfall varies tremendously from one point to another. And beyond a small region, all bets are off.

"If you have a very successful cloud seeding program, that is producing, maybe 7% or 9% or even 10% more precipitation over a mountain range, downstream of that mountain range, the impact may be 1%," French said. "But it is a really difficult number to get our arms around."

French added that scientists should be careful to not overpromise what cloud seeding can accomplish.

"If the promise is that cloud seeding is going to eliminate droughts even on a local level, the answer is no, it can't live up to that. There's no scientific evidence," French said. But if targeted properly, it can moderately increase natural precipitation, he said.

Cloud seeding on the Tibetan Plateau

Despite cloud seeding's limitations, China has established weather modification bureaus across the country.

"They have a whole campus of people who are working in the weather modification field," Rauber said. "They have a fleet of aircraft — they're on a whole different scale than anything that goes on anywhere else in the world."

Chinese researchers claim that the country has made huge strides in its weather modification. Reports from state-owned media suggested that weather modification efforts increased precipitation by 168 billion tons between 2020 and 2025, and that in experiments, just one cup of cloud seeding material generated 30 Olympic-size pools' worth of precipitation over an area the size of Yellowstone National Park in the arid Xinjiang region.

These weather modifications have also been conducted on the Tibetan Plateau. The vitally important region, nicknamed Asia's Water Tower, is the source of multiple major rivers that supply water to nearly 2 billion people across Asia. But the plateau is facing increasing desertification and glacier loss due to human-caused climate change and other human activities such as overgrazing animals.

China's cloud-seeding ambitions for the plateau reached a whole new level in 2018, when authorities announced the Sky River project. The controversial project aimed to use cloud seeding to create precipitation across a 620,000-square-mile (1.6 million square kilometers) area in the Tibetan Plateau — about the size of Alaska — to divert Indian monsoon rains above the Yangtze River basin and channel the water to the Yellow River basin. From the Tibetan Plateau, the Yellow River flows north and east into the parched northern regions of China, while the Yangtze flows south and east into less-water-scarce areas. The initial plan claimed the project would transfer up to 7% of the country's total annual water consumption, reports noted.

The atmospheric channel was intended to be part of the country's colossal South-North Water Transfer Project, circumventing the region's challenging terrain by moving the water in the sky instead.

The plan, due to be completed in 2025, called for seeding this atmospheric river using tens of thousands of silver iodide-burning chambers on the ground, linked to a series of meteorological satellites that would analyze the weather conditions. Information from these satellites would automatically trigger the burners when the conditions were suitable for cloud seeding.

However, the plan was immediately met with a barrage of criticism from scientists in China. In a translated statement, Hancheng Lu, a professor at the National University of Defense Technology's School of Meteorology and Oceanography, called the project "an absurd and fantastical project with neither scientific basis nor technological feasibility."

The project was unworkable because it is not possible to convert all atmospheric moisture to rain, or to channel moisture in this way, Yeh said.

As of 2022, researchers were working on a significantly scaled-back version of the project. However, China's official channels have gone dark on the topic.

China's newly announced five-year plan, which sets out the policy direction of the country from 2026 to 2030, references weather modification enhancements but doesn't mention this specific project, according to reports from Chinese media.

This has led experts to speculate that the project was quietly canceled.

"When one goes to China and asks atmospheric scientists not involved in it about it, they just sort of laugh, and it's like this embarrassing incident," Yeh said. "It was never possible."

To see whether the project was still going forward, Live Science reached out to several researchers involved in the project, but they did not reply by the time of publication.

The fears of cloud seeding

The pursuit of the Sky River project — despite the low likelihood that it would ever work —— has caused significant alarm in neighboring countries. India relies on the monsoon rains and rivers such as the Brahmaputra, which starts in the Tibetan Plateau before flowing through India and Bangladesh on a 1,800-mile-long (2,900 km) route to the sea. There are also suspicions about cloud seeding being used to cause flooding across borders.

Many of these fears are overblown, experts told Live Science. Rauber noted that cloud seeding can't influence weather enough to reduce water in a wider weather system.

"This is always a question of, 'Are you robbing Peter to pay Paul?'" Rauber said. But the amount of water in storm clouds "is way greater than anything cloud seeding is going to extract."

Climate geoengineering

China's all-in pursuit of technologies like cloud seeding — even on the internationally important and politically sensitive Tibetan Plateau — has raised concerns that China is prepared to go to extreme lengths to engineer its way out of its problems — even when it involves scientifically dubious geoengineering projects with massive risks.

China's determination to address climate change and investment in megaprojects that attempt to engineer the natural world, such as the upcoming Motuo megadam in Tibet and colossal tree-planting projects in northern China, are also signs that China is moving toward larger climate geoengineering, experts have suggested.

This may even involve attempts to change how much sunlight reaches the planet's surface, experts have speculated.

"Given similarities, weather modification could serve as a means to incrementally build legitimacy for solar radiation management, in China and beyond, which may ultimately make it possible to deploy it," researchers argued in a 2019 study.

However, there's no evidence that China is currently pursuing such climate modifications, and it's unlikely they'd do so without some cooperation or buy-in from other countries, researchers wrote in 2023.

"It is currently unlikely that China would deploy SRM unilaterally. But its weather modification programme does demonstrate the country's willingness and capability to undertake large-scale atmospheric intervention projects," the study authors wrote.

Solar radiation modification aims to release particles such as sulfur high into the atmosphere to reflect sunlight back into space in an effort to limit global heating. It is being increasingly investigated by researchers and governments. The controversial concept has never been demonstrated on a large scale, and studies suggest it has many potential risks, such as the potential to shut down monsoon rains.

Scientists have already seen some of the negative impacts of natural solar radiation modification, for instance in the aftermath of large volcanic eruptions, such as Tambora in 1815. "It led to the suppression of the monsoon, because the monsoon is driven by heating from the sun, Rauber said. "And that cuts back on rain in tropical areas, and that causes droughts, which causes all sorts of diseases and can lead to mass starvation," he added.

"The evidence from nature is don't mess with Mother Nature," he said.

Towering 14,800 feet (4,500 meters) above sea level, the Tibetan Plateau, with its frigid temperatures and glacier-covered mountains, is the source of most of Asia's major rivers — the Yellow, Yangtze, Yarlung Tsangpo (Brahmaputra) and Lancang (Mekong) — which supply almost 2 billion people downstream, including the two most populous countries on Earth: China and India.

For that reason, the Tibetan Plateau is often called the roof of the world, the third pole and Asia's water tower.

But in recent years, the region has also earned a new moniker: Asia's power tower, thanks to its huge, untapped potential for generating hydropower.

Faced with water scarcity in densely populated, industrialized and irrigated regions of China; an insatiable need for energy; and a drive to eliminate fossil fuel use, Chinese authorities are pursuing a number of hydropower projects in the region. Together, they will not only tap the region's vast power potential but also attain unprecedented levels of control over vital water sources its neighbors rely on.

The government's flagship project is the Motuo (also called Medog) megadam project on the Yarlung Tsangpo River. The project officially started construction in July 2025, and its costs are staggering — estimated at up to $168 billion.

The megadam is slated to be completed in less than a decade and will dwarf all other hydroelectric projects in the world with its estimated 300 terawatts of annual power output — three times the output of the Three Gorges Dam on the Yangtze, the world's current largest dam, and more than the entire U.S. produced in net hydropower in 2024.

The dam is just part of a bigger Chinese initiative to transform its environment.

"You have a modern, powerful China who is in a way very, very confident of taming nature," said Tenzin Norgay, a researcher at the International Campaign for Tibet (ICT), a nongovernmental organization that works to promote human rights and democratic freedoms for the people of Tibet. The ICT is closely monitoring dam building in the region. "That's literally what they are trying to do, right?" Norgay told Live Science. "They are trying to tame nature."

But the massive project comes with huge risks for both people in Tibet and the hundreds of millions of people in countries downstream, including those in India and Bangladesh, experts told Live Science.

"Controlling [the] nature of the water or the river itself is a danger for the entire Himalayan belt particularly for countries like India, Bangladesh and to some extent also Nepal," said Jagannath Panda, head of the Stockholm Center for South Asian and Indo-Pacific Affairs at the Institute for Security and Development Policy, told Live Science.

Project of the century

China is the world leader in dam building. The government has constructed around 98,000 dams and reservoirs across China, including 40% of the world's largest dams, and numerous dams outside the country as part of its Belt and Road Initiative. But this new project is different.

"There's nothing on this scale, and nothing close to it," Brian Eyler, director of the Southeast Asia and energy, water and sustainability programs at the Stimson Center think tank in Washington, D.C., told Live Science.

The project will utilize Tibet's unique geography to full effect. In the project area, the Yarlung Tsangpo River (known as the Brahmaputra in India and Jamuna in Bangladesh) flows through the world's deepest canyon, called the Yarlung Tsangpo Grand Canyon, and rapidly drops around the horseshoe-shaped "Great Bend," before continuing to flow down and into India and, eventually, Bangladesh.

The project will dam the upper section of the river and divert the water through a series of tunnels that will be cut through the 25,500-foot-tall (7,800 m) Mount Namcha Barwa, before returning the water to a lower section of the river, circumventing the Great Bend. The project will likely have five dams in total, with hydropower stations inside the tunnels. The water will drop 6,500 feet (2,000 m) within 30 miles (50 kilometers) of tunnels, thereby generating a huge amount of hydroelectric power.

"It's really incredible that this type of project can be built," Eyler added.

The cascading dam system will require huge amounts of water to run effectively, meaning there will likely be a reservoir at the beginning, and so during the dry season the Great Bend will effectively run dry, Eyler said.

Mega project, mega challenges

Building such a large project in this region is fraught with risk, experts noted.

The Tibetan Plateau is one of the most seismically active regions in the world, driven by the ongoing collision between the Indian and Eurasian plates. The region has recently been rocked by several major earthquakes, including the 7.1 magnitude Dingri quake in January 2025, which damaged five dams in the region, and the 7.8 magnitude Nepal earthquake in 2015 that damaged a fifth of the country's hydropower capacity.

Parts of the Himalayas are unsuitable for dam construction due to the high risk of seismicity and its effects, researchers have warned, noting that dams are particularly vulnerable to earthquake-induced landslides.

In addition to naturally caused earthquakes, large-scale projects involving land excavation, tunneling and water redirection have the potential to trigger seismic activity, while the creation of reservoirs has been strongly linked with earthquakes in China.

Other natural disasters could endanger people beyond Tibet.

Glacial lakes — bodies of water created by melting glaciers and permafrost — can pose a problem to people downstream if they suddenly release their water and overwhelm dams. This scenario happened in northeastern India in 2023, causing a large, newly built dam to catastrophically fail, killing at least 46 people and impacting 88,000 more.

Climate change is accelerating glacial melt, meaning the threat will only increase as the region's glaciers are further destabilized, experts have warned.

Climate change could also render the dam obsolete sooner than anticipated. Dams typically function for around 70 to 100 years, experts told Live Science. Water levels in the Yarlung Tsangpo are expected to peak in 2060, so when water levels start to fall after that, the whole dam could become useless during dry periods because the water level will be too low to produce hydropower — a status known as minimum pool elevation. Water levels could even drop so low that they can't pass through a dam. This situation, known as "dead pool," is already an issue for some dams on the Colorado River.

The project will also affect Tibetans. "From our viewpoint, displacement of people around that and submergence of cultural sites" are the biggest issues, Norgay told Live Science.

However, the area is sparsely populated, and the reservoirs needed will almost certainly not be as big or as deep as reservoirs of other megadams. So although there will be an impact, it won't be on the scale of the 1.3 million people displaced by the Three Gorges Dam, Eyler noted.

Downstream impacts

Even without a climate-driven reduction in water, the river's flow will be transformed. To ensure a continuous, controlled flow of water through the dam, authorities typically fill reservoirs during the wet season and release water during the dry season. While this ensures that the hydroelectric dam can function, it inevitably impacts the river's natural flow and has knock-on effects for communities downstream. The filling-and-releasing process also raises the possibility of the upstream country — namely, China — "turning off the taps" to benefit the dam at the expense of other downstream water users.

"If a dam operator has an opportunity to take water during a time of drought, they're going to take at the expense of downstream users," Eyler said. "We've seen this happen in the Mekong, where the downstream was suffering drought, but China still filled its reservoirs," worsening drought in 2019 in Thailand, Cambodia and Vietnam.

In the case of the Motuo megadam, any change in water flow will affect India and Bangladesh. The Brahmaputra flows for about 1,800 miles (2,900 km), eventually joining with the massive Ganges River, and is a vital source of water and natural fertilizer for more than 130 million people. It provides India with 30% of its freshwater reserves, while Bangladesh is heavily dependent on the river to support its irrigated agriculture.

"The greatest ecological and environmental impact will be related to sediment flow. The Great Bend itself is a rich provider of sediment to the downstream [countries]," Eyler said. "Sediment is important for agricultural production. It's a very inexpensive natural fertilizer."

Sediment from the Brahmaputra is integral to building up the Ganges-Brahmaputra delta, helping the low-lying region stay above rapidly rising sea levels. The delta is home to nearly 200 million people — including in the megacity of Kolkata — and is considered one of the places most at risk from sea level rise. The Yarlung Tsangpo in Chinese-controlled territory provides up to 50% of the river's downstream sediment flow.

"River deltas are built by sediment flows pushing land out into the ocean year after year," Eyler said. "And either the dams themselves or the lack of flow within the Great Bend will cause a great reduction in the amount of sediment coming down."

The potential reduction in sediment could threaten food security in the downstream countries.

"Millions of people's lives are dependent on this river," Norgay noted.

India also plans to build major dams on the waterway, Norgay added, which could itself have negative impacts downstream.

Not true "green power"

The Motuo megaproject is just one of several new dams planned or under construction in Tibet; the ICT counts at least 193 dams in the region that have been planned or built since 2000, when China embarked on a policy of expanding infrastructure projects in the region.

The surge in hydropower projects is intended to help China move away from fossil fuels, experts told Live Science.

"It is part of a more widespread strategic vision where China is trying to become more sustainable," said Tom Harper, a lecturer in international relations at the University of East London who specializes in China.

One key to China's strategic vision is phasing out coal.

"This dam has been described as the coal killer. There are numerous coal plants that can be taken offline as a result of this and retired permanently," Eyler said. "When you bring in China's carbon emissions reduction goals, through 2050, the dam makes a lot of sense."

However, hydropower is not as sustainable as it's often portrayed, and it has environmental impacts. Large dams can cause severe environmental damage to river ecosystems. The creation of reservoirs can also release greenhouse gases as trees and plants are covered with water and subsequently rot, though this is less of an issue in cold places like the Tibetan Plateau, Darrin Magee, a hydropower expert at Western Washington University, told Live Science.

Though the authorities claim there will be no significant environmental impact from the dam, that's hard to believe, Eyler said. "Typically, the larger the project, the greater the impacts, and this is the largest dam system ever created."

Scientists also question the necessity of using hydropower to meet sustainability goals, when the Tibetan Plateau has huge, untapped wind and solar power potential.

Experts had different theories of what China would do with such a vast amount of energy generated in such a remote and sparsely populated area.

"There's certainly no need for it, right now or in the foreseeable future in the area where Motuo Dam is sited," Magee said. "But China solves that problem by building ultra-high voltage DC transmission lines at, at a rate that no one in the world is matching." These high-voltage lines transport energy from the west to east, bypassing local grid networks, he added.

Norgay and Panda both think Tibetans are unlikely to benefit.

The power will likely be moved east to power Chinese industry, they said, and it could also align with the political goal of further integrating Tibet into China, Panda added.

Eyler, meanwhile, thinks it will be used to power data centers in Tibet, "which can be built around the super dam, in a naturally cool and cold environment."

Better cooperation needed

There is little official information about the megadam for Chinese authorities, Eyler noted, and the lack of transparency is fueling fears. For example, Indian politicians have expressed concern that the megadam will give China full control of the river, and that China could potentially weaponize ‬the river by deliberately reducing the flow of water or by releasing large amounts of water in one go, thereby devastating downstream communities. Some researchers have called this possibility a "water bomb."

Eyler, however, said that this risk is low, and that the dam system's design means it cannot hold back enough water to cut off supplies downstream.

"I don't think that there's some type of nefarious plot out there from Beijing to bring these countries to heel by controlling the upstream of the rivers," he said. "China's top priority is to develop its economy, bring stability to the country, and building large dams on rivers is one way to do that."

The expansion of hydropower is a key aim for China in the next few years, as the country embarks on the 15th five-year plan from 2026. With shared water resources in short supply, better cooperation among neighboring countries is vital, experts said. Yet China and India share only limited data, Panda noted.

Some of the concerns from neighbors could be mitigated by better communication between stakeholders, Magee said. "Be more transparent with the data, bring more voices into the conversation, have some realistic assessment of both need for the project and the impacts."

But even with better communication, the megadam and other upcoming dams means China will still largely control the region's water resources due to its upstream position.

"This is a dam project which actually gives China the upper hand," Panda said.

On March 10, the U.S. and Israel bombed Iran's capital of Tehran so heavily, one resident described the city as "the last stop before hell."

Buildings shook and windows shattered as missiles struck oil and weapons development facilities in Tehran, which is home to nearly 10 million people. The World Health Organization urged Iranians to stay indoors as acidic rain blackened with soot and toxic compounds showered residential areas.

Since the start of the war on Feb. 28, U.S.-Israeli forces have launched thousands of attacks around Iran, destroying homes, schools, hospitals and other key civilian infrastructure. Almost 3,500 Iranians have died, and more than 26,500 have been injured.

But as bombardments continue, another emergency is spiraling in the background. The war has overshadowed a dire water crisis in Iran that, among other things, caused Tehran to nearly run out of water late last year.

In November 2025, four of Tehran's five water reservoirs were 88% empty, and the fifth, Amir Kabir, was 92% empty. The city faced its worst water crisis in six decades, with taps going dry in some districts and desperation paving the way for deadly protests in December 2025 and January 2026.

Iran has been in a drought for more than five years, and water shortages caused by decades of overreliance on agriculture, broken water policies and hostility toward the rest of the world have gripped the country. Lakes, rivers and wetlands are drying up across most provinces. And in the northeastern city of Mashhad, which is home to about 3.5 million people, water levels in November fell to below 3% of storage capacity.

To make up for surface water losses, Iranians are pumping groundwater, but those reserves are also running dry.

"Iran, for the foreseeable future, will have many, many other issues that need to be sorted out, and I'm afraid the environment will not be at the top of priorities," Nima Shokri, executive co-director of the United Nations University Hub on Engineering to Face Climate Change and head of the Institute of Geo-Hydroinformatics at the Hamburg University of Technology in Germany, told Live Science.

But the water crisis hasn't gone away, and the war will likely worsen the problem, experts told us. Iran can't afford to neglect its water issues any longer, Shokri said.

"If the priority shifts in a meaningful way, if there is structural change, if the environment becomes a top priority, there are things that can be done," Shokri said. "Iran has no choice. Things will collapse unless there is meaningful structural change."

A water-bankrupt country

Tehran's water shortages last year persisted for months. Already in July 2025, officials warned that the city was weeks away from "day zero," when water supplies drop so low that taps run dry, forcing residents to queue for rations or buy bottled water. In August, faced with temperatures of up to 122 degrees Fahrenheit (50 degrees Celsius) that increased the demand for energy and water, authorities closed government offices, banks, schools and universities for several days.

By November, Iran's president, Masoud Pezeshkian, had revived a long-debated plan to move Tehran to the country's wetter south due to water shortages. This time, he claimed that relocating the city was "no longer optional." His government said the Makran region, which stretches across Iran's south from the Strait of Hormuz to Pakistan, could host a new city. But the government didn't release a detailed plan or address the relocation's cost, which analysts say could be more than $100 billion.

Pezeshkian's comments highlight the lack of a national strategy to deal with water scarcity, said Eric Lob, an associate professor in the Department of Politics and International Relations at Florida International University and a scholar in the Carnegie Middle East Program. Instead of implementing durable solutions, the Iranian regime has relied on water cutoffs, fines and temporary fixes to save water, which is like "putting a Band-Aid on gaping wounds," Lob told Live Science.

Government officials and experts dismissed Pezeshkian's claims as unserious, and they weren't wrong, Lob said. "Even if you were to do that, you still don't have a national plan in place, and you still only have these reactive, stopgap measures that are not going to bring forth any type of sustainable, serious plan."

But the fact that Iran's government would consider displacing 10 million people also shows the gravity of the country's water crisis, which has reached "water bankruptcy" levels — where the damage to some water systems is irreversible on human timescales. Nationwide shortages last year sparked protests in the central Isfahan province in March and contributed to country-wide unrest that the regime brutally suppressed in January 2026 by killing up to 30,000 people.

Last year worsened a historic drought, with Iran receiving about 40% less rainfall than the long-term average, and climate change is predicted to exacerbate the region's aridity. However, water bankruptcy in the country is not driven solely by climate change or short-term droughts. Rather, it stems from decades of broken water governance and aggressive policies that the regime has pursued at the expense of the environment and the Iranian people, Shokri told Live Science.

"It is not 'mismanagement,' because mismanagement is a shallow representation of the issue," Shokri said. "It's a high-level strategic error."

Self-sufficiency at any cost

Since the Iranian Revolution in 1979, the country's ruling elites have placed a great emphasis on independence and self-sufficiency, particularly in food production. In a commentary published March 13 in the journal Nature Sustainability, Shokri and his colleagues argue that this stance and Iran's post-revolution geopolitical aims caused its water crisis.

Agriculture in Iran has mushroomed over the past four decades, despite the country's arid climate and limited water availability. The regime promoted uncapped groundwater extraction to grow crops such as wheat, rice, barley and sugar cane for domestic consumption, while thirstier crops, such as dates and pistachios, simultaneously ballooned to feed exports. Officials invested only minimally in water reuse, leak reduction and monitoring infrastructure, and they heavily subsidized energy, which farmers used to drill more wells and pump more groundwater.

Additionally, the government funded hundreds of dams to store water, but some of the rivers on which they were built are too small to sustain reservoirs. "Oftentimes, the dams don't have enough water, and then also are causing some sort of evaporation, because there's just water sitting in there," Liz Saccoccia, a water security associate at the World Resources Institute in Washington, D.C., told Live Science.

Today, agriculture makes up roughly 90% of Iran's water use. As of 2024, 32 of the world's 50 most overpumped aquifers were in Iran. The number of wells to pump groundwater almost doubled from around 450,000 in 2000 to 800,000 in 2013, but the amount of water that was extracted from those wells in that period declined by 18% due to aquifers running dry. To get to the last drops, farmers have adopted submersible electric pumps that can reach below 160 feet (50 meters), but this has lowered groundwater tables so much that salt water has infiltrated aquifers. Farmers who use this water for irrigation now have saline soils, which limits their productivity.

Iran's food self-sufficiency policy has wavered over the past few decades, Lob said, and the country has imported wheat and other agricultural commodities, mostly from places like Russia and Brazil. But the pressure to operate water pumps has remained high, mainly because water use in Iran is extremely inefficient.

Inefficient water use is due to weak input from the Iranian government, which has spent resources on supporting militant and rebel groups around the Middle East, like Hezbollah in Lebanon and the Houthis in Yemen, rather than on developing sustainable water systems, Shokri said.

A lack of foreign investment and limited access to advanced water technologies from countries like the U.S. and Israel due to international sanctions have amplified the problem. Technologies such as big data analytics, artificial-intelligence-based irrigation scheduling and satellite-based water accounting could slash water use; instead, "the way they do the irrigation is completely outdated," Shokri said. For example, flood irrigation is used to grow rice in regions with high evaporation rates, such as Khuzestan province in southwest Iran, he said.

Other experts agreed that the Iranian regime is responsible for the country's water emergency and that global warming and economic sanctions are simply exacerbating an existing problem. "It's a crisis that's been years in the making, and I think that could be equally — if not more — attributed to government mismanagement, alongside these structural issues and externalities," Lob said.

The main flaw in Iran's water policy is that decisions are made on the national scale, rather than at a basin level, where authorities could plan economic activities based on the amount of water available, Saccoccia said. "In Iran, they have these political goals that do clash with the reality of the water resources that they have," she said.

From bad to worse

Freshwater depletion in Iran has helped trigger three other crises that threaten people's livelihoods, health and safety.

First, water shortages have impacted energy production, because water generates hydropower and cools power plants. In 2019, 10.6% of Iran's electricity came from dams, but in 2025, that percentage decreased to 3.3% due, at least in part, to record-low water levels in some reservoirs.

Power outages have become daily occurrences in cities like Tehran, where reservoirs have dropped well below capacity. Last summer, offices and shops were forced to close for hours each day due to electricity cuts that blocked internet access and appliances such as air conditioners and refrigerators. In the sweltering heat, water and electricity shortages made staying at home exceptionally difficult and left residents with few options but to buy bottled water to cool down.

The water crisis has also created dangerous air pollution through dust that blows off dry lake beds and riverbeds, Shokri said. Satellite images show that bodies of water across Iran have shrunk dramatically over the past 40 years, exposing sediments that are easily picked up by wind and can travel hundreds of miles. For example, northwest Iran's Lake Urmia — once the largest saltwater lake in the Middle East — has almost completely dried up and, in turn, increased harmful particle pollution in the region.

The third crisis born out of Iran's water emergency is widespread subsidence, or sinking of the land surface over time. This is due to groundwater extraction, which compresses sediments by removing water volume from the ground, Jess Payne, a doctoral student at the University of Leeds Institute for Geophysics and Tectonics in the U.K., told Live Science.

In a 2025 study, Payne and her colleagues used radar data from satellites to show that more than 12,120 square miles (31,400 square kilometers) of Iran — an area roughly the size of Maryland — is subsiding faster than 0.39 inches (10 millimeters) per year. Some places are sinking much faster than that; for example, Rafsanjan, a city in central Iran, dropped by over a foot (34 centimeters) every year between 2014 and 2022.

About 77% of subsidence in Iran occurs in valleys dominated by agriculture, and it is especially acute in desert regions, where rainfall doesn't replenish aquifers, Payne said. "Only 3% of the subsidence we observe happens under urban areas," she noted. "That tells you that the drivers are going to be probably linked to agriculture, and that probably tells you it's groundwater extraction."

But agriculture-driven subsidence affects urban areas, too. In cities like Shiraz, Isfahan and Yazd, uneven subsidence has caused huge cracks to form in buildings and roads. In September 2025, the threat was so severe that officials evacuated 40 schools in Isfahan, which has a population of about 2.2 million.

Subsidence may also increase the risk of earthquakes in Iran, which is a hotspot for seismic activity. Many faults are likely hidden beneath aquifers, and groundwater extraction could reduce the load on these faults, Payne said. "It won't cause an earthquake, but if that fault is close to failure, it might contribute to it happening sooner," she said.

In the longer term, subsidence is ominous because sediments can become so compact that they forever lose their ability to hold water. "Some subsidence is irreversible," Payne said. "You're permanently losing freshwater stores in the ground, and that's really problematic, particularly for arid regions like Iran."

Air strikes on critical infrastructure

The recent war has caused massive-scale destruction around the Persian Gulf and thousands of deaths. The conflict has eclipsed Iran's water concerns, but it will likely exacerbate shortages in the country over the coming months and years, experts said.

A major, imminent threat to Iran's water security is U.S.-Israeli forces' targeting of energy infrastructure, including fuel depots and oil refineries. Most water infrastructure cannot function without power, Auroop Ganguly, a distinguished professor of civil and environmental engineering at Northeastern University in Massachusetts, told Live Science in an email.

"Water and energy are inextricably linked," Ganguly said. "If energy infrastructure in Iran is targeted, water, wastewater and sewage treatment would be impacted, while electric wells, pumping stations, and water distribution networks would become inefficient or crippled."

Iran has also sustained heavy damage to a desalination plant on Qeshm Island, which officials said cut off water supplies to 30 villages. However, Iran relies on desalination for less than 3% of its water use, meaning strikes on desalination plants have a smaller impact on water security in Iran than they do in Gulf states like Bahrain and Kuwait, which depend on desalination for 90% of their drinking water, Ganguly said.

Desalination serves a handful of coastal cities and islands in Iran, including Bandar Abbas and Kharg Island. "Targeting desalination plants in the critical coastal or island regions would have severe local effects, but targeting larger dams or power supplies could result in broader nation-wide impacts," Ganguly said.

The air strikes have also caused other environmental problems.

The war has unleashed deadly air pollution from burning oil and gas facilities that will worsen existing air quality issues in Iran and fuel both immediate and long-term health crises, Frederick Otu-Larbi, a lecturer and environmental scientist at Ghana's University of Energy and Natural Resources and Lancaster University in the U.K., told Live Science.

"For several days or weeks at the beginning of the war, you could literally see black smoke covering most of Tehran, and Iran, and the region, from burning oil and infrastructure," said Otu-Larbi, who recently published an analysis of greenhouse gas emissions created in the first few weeks of the war. "We're looking at a whole country's total annual emissions being pumped out into the atmosphere in a space of 30, 40 days. All of that air pollution from burning oil [and] burning infrastructure is likely to cause some immediate problems, but also long-term health complications for the population."

"Chaos" ahead

The war will likely worsen Iran's water crisis by diverting its resources to reconstruction rather than the environment and by further isolating the country from neighbors and others that have more advanced water technologies.

Solving Iran's water emergency requires creating alternative economic opportunities for farmers and boosting water use efficiency across the country, yet the war will likely strengthen the regime's goal of self-sufficiency and entrench its geopolitical stance, Lob said. "If anything, this war has hardened the regime," he said, adding that this makes it extremely unlikely that Iran could obtain advanced water and desalination tech from countries like the U.S. and Israel in time to stave off a water catastrophe. These technologies would not make up for decades of water depletion, but they could alleviate pressure on Iran's natural systems and help the country save water.

Many farmers cannot move into other sectors, because economic sanctions and Iran's isolation from international trade and investment restrict what jobs are available. Reconstruction after the war could cost tens, if not hundreds, of billions of dollars, further limiting economic growth and job opportunities, which would bake in more water depletion through continued reliance on agriculture, Lob said.

Iran wants to increase its desalination capacity to mitigate water shortages, Ganguly said, and it is building pipelines to carry desalinated water inland to cities like Isfahan. But desalination plants are expensive to operate, and they take up to six years to erect, Naser Alsayed, an environmental policy expert for the Gulf region at the University of London's School of Oriental and African Studies, told Live Science. Iran is a mountainous country, meaning transporting desalinated water inland will be difficult, Alsayed added.

The longer the war drags on and the more Iran delays rethinking its water management and national priorities, the worse the water situation will get. Nature takes 20 to 30 years, on average, to recover from warfare, a 2025 study found. Water supplies could be contaminated with heavy metals, oil and other toxic pollutants for decades to come. This potential contamination, together with estimates that population growth will drive up Iran's water demand by 30% by 2050, suggests the country will face more severe water shortages in the future.

"If there is no structural change in the governance, things will only get worse," Shokri said. "The environmental system will collapse."

Water-shortage protests that have been building over the past few years indicate that Iran's society could break down if the regime doesn't act soon to preserve water, Alsayed said. "If people don't even have access to essential resources, then it's going to be chaotic — very chaotic," he said.

More revolts over basic access to water and electricity could lead to more violent crackdown by the regime and avoidable deaths, as well as emigration from the country. "The regime will find it hard to contain this; that's for sure," Alsayed said.

The solutions are numerous and obvious, Shokri said. "How about you heavily invest in wastewater treatment? How about you modernize your irrigation system? How about you start normal interactions with the world? How about stabilizing your currency? How about you use money to strengthen infrastructure? How about fixing your water transportation system?"

At the very least, Iran should integrate water use efficiency measures into its postwar reconstruction, Lob said. "We're looking at a country that's been devastated by war, so the situation is pretty bleak, and they're dealing with an economy that's in shambles," he said. Nevertheless, there could be an opportunity to build back better when it comes to water infrastructure. "One can hope."

My first clue that something was wrong was the smell of cotton candy.

It was March 15, 2016, and I was changing my toddler's sheets because he had soaked through his diaper. The bed was wet, but it didn't smell like urine. Other details I'd brushed off came to mind, and I began to worry.

He'd lost his chubby baby cheeks. He'd been finishing his morning sippy cup of water, rather than flipping it over and cackling in glee. And he seemed a bit more tired.

I called the pediatrician, who tested my son's urine. It showed dangerously high sugar levels. After a harrowing trip to the hospital, we had a diagnosis. My son had type 1 diabetes, like 2 million children and adults in the U.S., which meant his rogue immune system had destroyed almost all of the insulin-making beta cells in his body. He was 1 year old.

A decade later, he is a thriving sixth grader.

But in his own words, "diabetes really sucks."

"There are times when I can't eat because my blood sugar is high. Sometimes, when I'm really low, I'm scared that I'm going to die," he told me.

If he lives to old age, he will have changed the sensor that measures his blood sugar and the "site" — the tube that delivers insulin — tens of thousands of times, will have spent more than half a million hours with high or low blood sugar, and will have incurred millions of dollars of medical care. Getting off this blood sugar roller coaster for good would dramatically reshape his future.

"I could get more sleep. And I wouldn't have to do any site or sensor changes, or get shots," my son said. "It would be a better life."

There's a running joke in the type 1 diabetes community that a cure is "only five years away" — and has been for at least 50 years.

But there may be a kernel of truth in it now.

The first treatments to replace destroyed beta cells are in late-stage trials that will likely conclude in the next few years. And scientists are developing innovative ways to keep those transplanted cells alive without harsh side effects.

Many hurdles remain before such treatments can be widely used in kids with the disease. But a picture of a post-type-1-diabetes (T1D) world is emerging. Not too far off, those who have dangerously erratic blood sugars or who already need to suppress their immune systems for other reasons may be freed of the disease with stem cell transplants. And after that, otherwise healthy adults and kids may receive stem cell transplants.

"Cell replacement therapies are our flagship program because we believe it's imminent to provide a functional cure for people with type 1 diabetes," meaning the symptoms of the disease are eliminated, said Dr. Sanjoy Dutta, chief scientific officer at Breakthrough T1D, a nonprofit that supports research and advocacy to prevent, treat and cure the disease.

From death sentence to chronic disease

My son's most noticeable symptom — the sweet smell of his urine — was a textbook sign of the disease. "Diabetes mellitus" (the full name of the disease) means "sweet siphon" in Greek, and its symptoms were first described 3,500 years ago in ancient Egypt.

We now know type 1 diabetes occurs when the immune system mistakenly destroys insulin-producing beta cells in the pancreas, which are clustered in cells called islets.

As a result, the body can't move glucose into tissue. Instead, glucose piles up in the blood, and the kidneys go into overdrive as they try to flush it out. Meanwhile, muscle and fat tissue starves. Without glucose for fuel, the body breaks down fat for energy and turns it into compounds called ketones. This water-intensive process dehydrates the body even further, and these ketones accumulate in the blood, turning it acidic.

A person with untreated type 1 diabetes will be ravenous and thirsty. As they eat and drink, and pee and pee, their body will slowly devour itself as their fat and muscle melt away. Without treatment, a person with type 1 diabetes will invariably go into a coma and die.

The solution to this problem is insulin. In 1921, Frederick Banting and Charles Best isolated insulin from pancreatic islet cells in dogs and eventually devised a way to make insulin from the pancreas of a cow.

Science had transformed type 1 diabetes from a death sentence to a chronic disease — an accomplishment that earned Banting the 1923 Nobel Prize in medicine.

We've come a long way since then. People now routinely live decades after their diagnosis. Scientists have made faster-acting insulin, designed insulin pumps that automate the delivery of the drug, and created continuous glucose monitors that reveal blood sugar levels up to every minute. These innovations improve average blood sugar levels and reduce complications.

But none of these advancements has fundamentally changed the relentless nature of the disease. Injected insulin doesn't work as fast as the version the body produces. And because dozens of factors — including diet, exercise, stress, illness and puberty — can affect how much insulin the body needs, patients are always at risk of under- or overdosing insulin.

"This is a disease that requires 24/7 superhuman management," Dutta told Live Science.

People who inject insulin typically face hours of high and low blood sugar every day. Decades of high blood sugar can destroy the kidneys, damage nerves, and cause blindness and heart attacks. Critically low blood sugar can kill within hours, as the brain is starved of energy.

Despite the medical advances, type 1 diabetes still shortens life expectancy by more than a decade, on average.

Replacing lost cells

The reductions in both the length and quality of life have pushed scientists to continue seeking a cure for diabetes. Merely managing the disease isn't enough.

Since the 1990s, more than 1,000 people worldwide have received transplants of islet cells from cadavers; in the U.S., this remains an experimental procedure, so it is done only in clinical trials. Some of the islet-cell recipients have been cured of the disease — at least for a time.

But the donor pancreases the islet cells are taken from are in very short supply. And cells can be damaged, either before or during harvesting, so their quality is often poor. As a result, cadaveric islet cell transplants aren't very effective; five years after transplantation, only 1 in 10 people who receive them don't need to inject insulin.

Another challenge is that recipients must take immune-suppressing drugs for life. As a result, cadaveric transplants are mostly offered to people who already need another type of transplant, such as a kidney transplant.

For islet transplants to represent a viable cure, they need to be easy to mass-produce in consistently high quality, and scientists must prevent the immune system from destroying the islet cells.

We are getting very close to that first goal: producing high-quality beta cells from scratch.

From embryonic cell to beta cell

In the past several years, scientists have made incredible progress in deciphering the chemical symphony that directs pluripotent embryonic stem cells — cells that can become any type of cell in the body — to morph into islets.

"Even five years ago, I wouldn't have predicted we were as far along as we are," said Lori Sussel, director of basic and translational research at the Barbara Davis Center for Diabetes at the University of Colorado Anschutz Medical Campus, who does basic islet research using mouse and human stem cell models.

Some treatments built on this research are already in late-stage development. For instance, Vertex Pharmaceuticals, a Boston-based biotech company, reported in 2025 in the The New England Journal of Medicine (NEJM) that 10 out of the 12 patients who received the company's embryonic stem-cell-derived product, called VX-880, do not have to take insulin a year out from their transplant. And all of them are making some insulin, Felicia Pagliuca, senior vice president for cell and genetic therapy research at Vertex, told Live Science.

(The two patients who are still insulin-dependent took steroids during treatment against study instructions. Steroids are known to raise blood sugar, which can damage beta cells, the authors of the NEJM study wrote.)

The trial started out with 14 participants, but two people died during it. However, the study authors, and presumably the safety monitoring boards for the trial, concluded that those deaths were unrelated to the treatment. Nonetheless, as with any potential treatment, more data is needed to see if further issues crop up among patients undergoing these transplants.

The company is doing just that, by testing VX-880 in a larger trial that will enroll 50 patients with hypoglycemia unawareness, meaning they can't physically feel when their blood sugar gets dangerously low. These patients face a higher baseline mortality risk, Sussel said. The trial will need a year's worth of data before it can be submitted to the Food and Drug Administration, Pagliuca said. But if those results are positive, that would mean a scalable source of reliable stem cells could be available for some people with type 1 diabetes.

Vertex isn't the only team working on reprogramming cells to become islets. Researchers in China have reprogrammed a woman's own fat cells to become beta cells. One advantage of this approach is that those cells aren't immediately recognized as foreign, because they come from the patient's own body. A downside, however, is that creating a treatment that is unique to each individual is likely to be cost-prohibitive as a broad-scale cure, Sussel said.

Immune invisibility cloak

Creating the stem cells is just the first step. Once they're in the body, they have to survive. That has historically required strong immunosuppressive drugs, just as other transplants do.

Immunosuppressive drugs — such as tacrolimus and sirolimus — leave people susceptible to infections and can damage the kidneys, and may increase the risk of cancer. Tacrolimus is also toxic to the transplanted islet cells themselves, said Dr. Piotr Witkowski, director of the Pancreas and Pancreatic Islets Transplantation Program at UChicago Medicine.

But the average kid with type 1 diabetes has decades of relatively healthy life ahead of them, so putting their heart and kidneys at risk with immunosuppressive drugs presents a higher risk than benefit.

"Immunosuppression is worse than having to take insulin every day," Sussel told Live Science. Thinking about it for my own son, I can't help but agree; diabetes is a huge hassle, but running the risk of fatal infections or higher cancer rates down the line to eliminate that hassle doesn't seem worth it.

Scientists are exploring several ways to solve the immune attack problem.

One is to make transplanted beta cells "invisible" to the immune system. To do that, one company is taking a page from pregnancy. During pregnancy, 50% of the genes — and, therefore, proteins — in the fetus come from the father, said Dr. Sonja Schrepfer, scientific founder of Seattle-based Sana Biotechnology (Hypoimmune Platform) and researcher-scientist at Cedars Sinai Medical Center in Los Angeles.

So Schrepfer's team studied which molecules help shield the fetus from immune attack to sustain the pregnancy.

They identified a few. Two molecules act as unique fingerprints present on the outside of every cell in the body. When the immune system detects a foreign fingerprint, it destroys the cells carrying it. During pregnancy, the fetus dials down how many of these fingerprints are made, especially in the placenta, Schrepfer said. So, to keep transplanted islet cells safe from the immune system, "we have to get rid of that fingerprint," Schrepfer told Live Science.

But that's just the first step in preventing rejection. Certain immune cells, called natural killer cells and macrophages, constantly patrol the body and will automatically kill cells without a fingerprint, she said.

However, Schrepfer's research showed that cells naturally produce a special molecule, at low levels, which essentially screams "Don't kill me." When cells overexpress this molecule, the macrophages and natural killer cells come close to transplanted cells but then drift away, leaving the cells alone, Schrepfer said.

So Sana Biotechnology has used the gene-editing cut-and-paste tool CRISPR, as well as a harmless virus, to genetically modify beta cells collected from a cadaver to lack the cell fingerprints and to overexpress the "don't kill me" molecule. These "hypoimmune" cells are, in theory, tolerated by the immune system.

In August 2025, Sana reported that it had treated the first patient with these cells earlier that year, and in March 2026 said the cells were still evading the immune system and making insulin . While larger safety and efficacy trials are still a ways off, the new results raise hopes that this approach will lead to stem-cell-derived beta cells that don't require immune suppression.

Vertex Pharmaceuticals is also investigating gene-edited stem cell transplants that can evade immune detection, but the company has not tested them in humans yet, Pagliuca said.

Gentler drugs

Other ways to prevent transplant rejection are under development. Unlike the gene-edited islet cells, these methods rely on a more established drug approval pathway.

One is a drug called tegoprubart, which was initially developed to prevent kidney transplant rejection. (It is also in trials as a treatment for delaying ALS progression.)

When immune cells first recognize a foreign protein, they need a second signal to mobilize the rest of the immune system to attack, Witkowski told Live Science. Tegoprubart silences that second "attack signal" by inactivating a molecule called CD40 ligand, he said.

In a study involving 12 patients, Witkowski and his colleagues transplanted cadaveric islet cells and gave the participants tegoprubart instead of the standard-of-care drug, tacrolimus. Tegoprubart led to greater beta cell survival than is typical with tacrolimus.

In March, Witkowski presented results at a diabetes conference showing that all 10 patients who received transplants more than four weeks earlier are so far are insulin-free and as of yet, none of the participants have experienced severe side effects or rejection.

Later this year, the team is also planning to start a 10-person trial with Breakthrough T1D to test tegoprubart in those with kidney dysfunction and T1D — a group that has historically been prevented from undergoing these transplants, Witkowski said.

Moreover, tegoprubart is just one of several gentler immune-suppressing protocols that are currently being studied, Dutta said. Others include antithymocyte globulin and teplizumab, a drug that has been demonstrated to prevent the progression of early-stage type 1 diabetes, he added.

Hope on the horizon

Type 1 diabetes research is littered with dozens of once-promising cures that ultimately fizzled, so it pays to be cautious about these new approaches.

Indeed, there are still some kinks to work out. For instance, when embryonic stem cells are converted to beta cells, most of the resulting cells do produce insulin, but some also make other hormones, such as glucagon, which raises blood sugar, and somatostatin, which inhibits growth hormone, Sussel said. Plus, the process for making these cells is laborious and time-consuming.

"Scaling it up is going to be very expensive, and so I think right now, that's probably one of the rate-limiting steps, is the expense," Sussel said.

And, of course, longer-term studies will need to pin down how long stem cell transplants will last and whether they need to be readministered periodically. There could also be side effects that take years or decades to emerge. As a parent, I'd have to be confident the deaths seen in the Vertex trial were completely unrelated to treatment before I'd be on board.

Everyone I spoke with was wary of pinning a time frame to a post-diabetes future.

"I would hate to say we're five years away," Sussel said. "What I would say is that there is a cure on the horizon."

But there was a palpable excitement among the researchers I spoke with, all of whom said breakthroughs are coming at a rapid pace.

A world where no one has type 1 diabetes is still a long way off. But little by little, a functional cure is coming for more and more people.

"Ideally all people receive it one day, but we're going to start small and then go big," Dutta said.

My son realizes he probably won't see a cure for a while. But knowing something is coming gives him hope. And that would make it all the more heartbreaking if these treatments don't pan out.

"When I imagine myself as an adult, I don't see myself with diabetes," he said.

"I'll be very disappointed if these treatments don't work."

There's a spa floating in the middle of Lake Erie. It has a sauna, a steam room and even a cubicle filled with snow. Upstairs, there are luxury lounges, a huge library, a curated art collection by notable artists, and a panoramic lecture theater with floor-to-ceiling windows. Passengers are busy dining, surrounded by sommeliers, in fine restaurants.

One deck below, there's a pristine, state-of-the-art laboratory full of high-tech equipment, and two multimillion-dollar submersibles can take passengers down 1,000 feet (300 meters). A team of scientists is sifting through water samples and analyzing them in real time, looking at the genetic fingerprints of plankton as it floats through the water.

The researchers on Viking's Octantis cruise ship are studying environmental DNA (eDNA) — bits of genetic material that float in the water, drift through the air, or linger in the soil. Every time a living creature passes through an environment, it sheds minuscule bits of its genetic material.

Scientists first noticed traces of this genetic material decades ago, but thanks to powerful sequencing techniques, they are now beginning to analyze eDNA to characterize food webs, reveal the locations of long-lost endangered species, and show if predators are lurking in areas where humans and wildlife are in conflict. But the technique has one problem: It generates so much data that researchers struggle to analyze it all. Now, scientists are working to combine artificial intelligence (AI) with cutting-edge sequencing to rapidly identify changes in the types and numbers of organisms in a given ecosystem. Eventually, that information could provide a real-time view of how the planet operates — and allow us to adapt to ecological changes more quickly.

"AI's going to be able to pull out [information] in a way that our other techniques just don't have the capabilities to," said Zachary Gold, research lead of the Ocean Molecular Ecology program at the National Oceanic and Atmospheric Administration's (NOAA) Pacific Marine Environmental Laboratory. "Quicker, better, faster data allows us to do things we've never dreamt of before," he told Live Science.

A treasure trove of environmental data

The term "environmental DNA," or "eDNA," was coined in the 1980s in a study describing a technique for getting DNA from a soil sample. But it wasn't until the 2000s that fast and accurate DNA sequencing machines became widely available and affordable, making eDNA analysis practical.

Next-generation sequencing (NGS) now allows scientists to analyze DNA incredibly quickly — the entire human genome can now be sequenced in just one day. For eDNA, NGS means thousands of species can be identified from a single water sample. The sequencing technology is highly advanced, but the ability to analyze and draw meaningful conclusions from it requires a huge amount of computing power and could take years of scientists' time.

The physical samples can take anywhere from a couple of days to a month to sequence, then once the sequences come back, many gigabytes of data must be downloaded and "cleaned" — that is, checked by a computer for mistakes, duplicates or formatting issues. Only then can validated datasets be analyzed.

It's that next step where AI could be transformative.

"Researchers can spend months looking through that data to try to understand and identify what are the most interesting and more powerful stories and assets that are coming out of this data, but the AI could do it, you know, in seconds," Gold said.

An army of floating laboratories

Viking began studying eDNA in part because of the pandemic. The company initially required guests to take daily polymerase chain reaction (PCR) tests for COVID-19, but once that requirement was phased out, the equipment on board its ship Octantis was repurposed to allow for real-time testing of water samples. The cruise company teamed up with NOAA in 2020, and scientists joined Viking's expedition to the Great Lakes in 2022.

Now, scientists aboard this 673-foot-long (205 m) cruise ship analyze phytoplankton in the waters they pass through, providing a snapshot of the ecosystem each time the ship visits the same regions. Compared with traditional scientific research expeditions, which are expensive and irregular, tourism vessels save time and money — cruise ships are going on these voyages anyway — and the food is a lot better, the team said.

In their floating lab, researchers working with Viking now sequence phytoplankton. "They are the key to life on Earth," said Allison Cusick, a researcher at the Scripps Institution of Oceanography at the University of California, San Diego, who works in one of Viking's ship laboratories to study eDNA in remote locations like Antarctica. Phytoplankton are the foundation of most marine food webs, and they produce about half the planet's oxygen via photosynthesis. The differences among phytoplankton species is mind-blowing — the diversity between two types can be greater than that between a human and a fungus, Cusick said.

Changes in the type of plankton in the water are key indicators of biodiversity and ocean health — shifts can ricochet up the food web, with potentially devastating consequences.

Using eDNA analysis to uncover evolutionary relationships between species and the different evolutionary paths they took — for example, when one arose and when specific genes were introduced — could help scientists predict how climate change will affect different species, said Benoit Morin, a supercomputer engineer at IFREMER (the French National Institute for Ocean Science and Technology).

"By looking at the past, we can try to understand the future," Morin told Live Science.

An "Enigma project" for eDNA

To be really powerful, projects like the Viking-NOAA collaboration will need to integrate artificial intelligence into eDNA analysis.

Already, AI is being used to find potentially new species from large data sets from camera traps and automated monitoring systems. It's also being used to rediscover lost species, including the critically endangered De Winton's golden mole (Cryptochloris wintoni), which, until it was traced using eDNA, hadn't been seen for over 80 years.

But for these efforts to reach their full potential, AI techniques will need to be refined and integrated into eDNA analysis.

Once scientists have collected an eDNA sample, they analyze it via bar coding, which can either look for a single species or organism or identify multiple species at once. The barcode is a small series of unique DNA sequences that are used to identify an organism by comparing it to an online reference database.

Letizia Lamperti, a mathematical engineer at the École Pratique des Hautes Études (Practical School of Advanced Studies) in France, is developing a machine learning system to use such bar coding to reveal the health of a given environment, based on the type and number of organisms within a sample. That information, in turn, could point to potential fixes.

For example, if there was an increase in toxin-producing phytoplankton in a water sample, it may be possible to pin those changes to agricultural runoff that's feeding the phytoplankton, Cusick said.

In 2023, Lamperti and her colleagues published a study showing that neural networks — multilayered machine learning algorithms that mimic the way the human brain filters and processes information — do a better job than other statistical methods of grouping closely related organisms based on their eDNA. But just like facial recognition technology, AI will likely be better at detecting abundant species, for which there is a lot of "training" data, but less effective at spotting rarer organisms.

Several other recent studies point to the promising potential for AI in eDNA research. For instance, one study found that AI can identify 90% of unknown species in a sample, even when there aren't similar sequences from closely related organisms to use for comparison.

If AI can fulfill its potential, the shift in how we understand the environment would be monumental. Cusick likened it to Alan Turing's decryption of the Germans' Enigma code during World War II. "That's going to be transformative," she told Live Science.

AI could identify newfound species on an unparalleled scale. Evolutionary relationships could be determined in the blink of an eye. Monitoring and planning for environmental changes could be transformed. For instance, by rapidly analyzing eDNA samples, AI could alert swimmers in real time to the presence of brain-eating amoebas or sharks in waterways, or forecast events like harmful algal blooms before they threaten public health — similar to how we get weather alerts on our phones now.

In theory, then, resources could be redirected quickly to resolve issues before they become a problem.

This goal is achievable, Gold said, but how long it will take will depend on the resources funneled into developing the AI to do so.

A dictionary of species

At the moment, AI is missing something important: organized volumes of good data for spotting key patterns. These data need to be put in one place as a reference database, or a dictionary of species, based on their DNA.

"We need the database of reference to perform the species identification," Lamperti told Live Science. "The problem is that we don't have it."

To identify species, AI needs to learn the key signatures, or barcodes, of individual and closely related species by training on reams and reams of data. But biodiversity datasets are not in publicly available repositories, and they're not in curated, standardized formats that can be fed into trained, bespoke AI systems. "eDNA is not AI-ready," Gold said.

In the U.S., around 40,000 eDNA samples have been collected in the past decade alone, Gold estimated, but a lot of it isn't accessible. It could be "in somebody's attic or the supplemental methods of someone's scientific paper," he said.

To draw useful conclusions to help us protect and manage the environment, AI needs to learn from a baseline database that captures biodiversity in the environments we're interested in. That's a herculean effort. "It's millions of dollars; it's tons of people's time," Gold said.

Morin is currently working on this task, but it's a slow and resource-intensive process. He and his colleagues are building a genetic "dictionary" through the ATLASea project, which aims to sequence the genomes of 4,500 marine species. This information will be deposited in an open-access database for the scientific community. IFREMER is now working with data infrastructure company NetApp to classify the mass of information being collected.

With money to develop the datasets, an AI eDNA tool could be ready "really fast," Gold said. "I have no doubt that what we're doing is not technologically difficult. It's just we're not resourcing it. If we really wanted to do this and mobilize at a scale, I have no doubt by the next Olympics in Los Angeles [in 2028], we could have the tools and resources and network set up and [be] ready to do this."

If investment and resources continue at their current pace, Gold estimated it will be a "slow trickle" and we'll get there in around 15 years. But he's optimistic the timescale could be faster. "A lot of the stuff isn't hard; it's just taking the existing tools that are already out there," Gold said. "We've just got to point the bike in the right direction."

Editor's note: This article was originally published on Feb. 25, 2025. It was previously updated to clarify that Viking, not external agencies, required guests to take daily PCR tests during the COVID-19 pandemic.

Between the 17th and 20th centuries, humans killed millions of whales for oil. They stripped their blubber, spinning the iconic creatures in the water and pulling off the fat in a huge spiral like the peel of an apple. The blubber was boiled into oil, then strained into barrels to be used in everything from oil lamps to industrial lubricants.

This was the bloody process that brought light to society.

"It is horrible," Charles Nordhoff wrote of his experience on a whaling vessel in 1895. "Yet old whalemen delight in it. The fetid smoke is incense to their nostrils. The filthy oil seems to them a glorious representative of prospective dollars and delights."

For over 100 years, the voracious hunger for whale blubber drove blue, humpback and North Atlantic right whales to the brink of extinction.

Now, commercial whaling is all but banned, whale blubber is used in just a handful of products, and whale populations have rebounded somewhat.

A similar sea change is coming for petroleum, though when and how it will play out is still incredibly hazy.

The best superforecasters, combined with machine learning, are only accurate at predicting geopolitical events up to a year in advance, Luke Kemp, research affiliate with the Centre for the Study of Existential Risk and at the University of Cambridge, told Live Science. At best, "we have general pictures we can paint."

But the general trends are clear. We've already transitioned much of our home energy use away from oil. And as climate change pushes us to accelerate that transition, we're developing new technologies that will help the world outgrow its oil dependence ever faster, experts say. In a few industries, like shipping and plastic, the decayed bones of long-dead animals will be the primary energy source for a long time to come.

But a post-oil world is coming.

"The whaling industry is a very good analogy," David MacDonald, a professor of petroleum geology at the University of Aberdeen in the U.K., told Live Science. At its peak, "The whaling industry was huge." But over the decades, "it was an inexorable decline," he said.

Origins of oil

Humans have been using oil for millennia. In fact, around 40,000 years ago, people in what is now Syria used bitumen — a byproduct of crude oil — to stick handles onto their tools. Fast-forward 35,000 years, and the Mesopotamians used the same sticky substance to waterproof their boats. The Babylonians used it to build the Hanging Gardens, and the Egyptians used it to embalm mummies.

In China, people burned crude oil and gas for heat and light as early as 500 B.C. By the fourth century A.D., they were drilling for these natural resources and transporting it via bamboo pipes.

But it wasn't until 1859, when Edwin "Colonel" Drake struck it big in Pennsylvania, that oil was sought at scale. With the same, albeit modernized drilling technique used in China more than 1,500 years earlier, Drake hit a reservoir 69.5 feet (21 meters) down, and the U.S. oil industry was born.

Crude oil, which is composed of simple strings of carbon and hydrogen, forms from the remains of animals and plants that sank to the bottom of swamps, lakes and oceans. Over millions of years, layers of sand and rock covered them, and intense heat and pressure turned these remains into oil and natural gas. They were then locked away in reservoirs — some close to the surface, others thousands of feet below — with gas sitting atop a lake of oil.

For the past 165 years, crude oil has transformed virtually every aspect of society.

If oil vanished tomorrow, global trade would break down as the shipping and aviation industries ground to a halt. Food security would be precarious, with no petroleum to fuel large-scale agriculture or packaging to keep food fresh. Medical care would be set back generations without the sterile equipment needed in hospitals. Renewable energy projects would be frozen without the components required to make solar panels or wind turbines.

Planes, trains, boats and automobiles

Our transition away from oil will be far gentler than that, of course. We've largely stopped using oil for electricity. In October last year, a report by the International Energy Authority found demand for oil will peak this decade.

The rise of electric vehicles (EVs) will usher in the next big drop in oil use.

Currently, road vehicles make up almost 50% of global crude oil use, according to a 2018 report by the International Energy Agency (IEA). But this percentage will plummet in the coming decades. It's estimated that EV sales will account for over two-thirds of the global market by 2030. If we are particularly aggressive in slashing fossil fuel emissions by three-quarters by 2050, the EV industry could be responsible "for more than half of the reduction in total oil demand," according to the BP Energy Outlook 2023, which forecasts future fuel use.

In 50 years, most of this car-driven oil usage could be eliminated.

Aviation also largely relies on oil for fuel. Planes last decades and cost tens of millions to hundreds of millions of dollars to build. But technology is moving fast in this sector. New aircraft are far more fuel efficient than aircraft were 40 years ago, and the industry is engaged in reaching net-zero emissions by 2050.

Sustainable aviation fuels (SAFs) will be key to ditching oil. These biofuels are derived from the raw materials used for industrial processes, including waste, biomass, cooking oil and animal fat waste. SAFs have the added benefit of being compatible with current aircraft engines, and they can be blended up to 50% with traditional jet fuel. Boeing plans to make all of its commercial aircraft capable of flying on SAFs by 2030. By 2050, if we aggressively cut carbon emissions, SAFs will account for between 30% and 45% of aviation fuels, BP estimates.

Shipping is a more stubborn problem. Ships run on oil. Like planes, they are wildly expensive to build, last decades and will be hard to phase out. Around 90% of world trade is carried out by the international shipping industry, with over 105,000 merchant ships currently sailing the oceans and accounting for around 5% of oil consumption today.

Without ships transporting goods all over Earth, "half the world would starve and the other half would freeze," according to the International Chamber of Shipping. The problem for this industry is, you can't just change the fuel.

Fredric Bauer, an associate senior lecturer at Lund University in Sweden, researches low-carbon innovation in energy and industrial systems. He's not convinced the shipping industry will be able to transition away from oil anytime soon. The International Maritime Organization published its first climate strategy in 2018 and has generally been "incredibly conservative" in shifting away from fossil fuels, Bauer said.

Hydrogen is a potential alternative fuel. Ships could be retrofitted with hydrogen fuel cells, but that strategy comes with problems. For example, to remain liquid, the fuel must be stored at extremely low temperatures. Its energy density is low, increasing the amount of storage required on each ship. Hydrogen is also extremely explosive.

Hydrogen-powered ships are still in their very early stages. The first ferries and small ships using this technology are being tested, but large, hydrogen-fueled oceanic cargo ships are still in the design phase.

Jay Apt, a professor at Carnegie Mellon University's Tepper School of Business and the Department of Engineering and Public Policy, told Live Science that shipping will likely be a voracious oil user for decades.

"If I was to look into the cloudy crystal ball, I would say that long-haul shipping would be one of the large-scale uses of petroleum that we would see 100 years from now," Apt said.

Plastic fantastic 

Single-use plastics are littering Earth in ever-increasing quantities. They take hundreds of years to degrade and then become microplastics, which are choking the ocean, littering the tops of mountains and congregating inside our bodies.

"The use of plastic is in many ways the more dangerous part of the oil industry, rather than the burning of hydrocarbons," MacDonald said. "If humanity disappeared from Earth tonight, in 1,000 years the levels of CO2 in the atmosphere will be back to normal — whatever normal is — but there would be plastic in the oceans and soils for millions of years."

Synthetic plastic is made from oil, and it is extremely cheap to produce.

Around 12% of the oil extracted today goes toward the petrochemical industry, which makes plastic and fertilizers, along with clothing, medical equipment, detergents and tires. And this number is set to grow: The Organization for Economic Co-operation and Development estimates that under current policies, the global use of plastics could triple by 2060.

Plastic is extremely useful because its density can be varied. We can try to move away from plastic in products like food packaging, but phasing out medical plastic is more challenging. Plastic is everywhere in hospitals, including in disposable syringes, IV bags, catheters, gloves and bed linens. It's not just that plastic is cheap, durable and malleable. It's also sterile, so helps curb the spread of infections.

"I couldn't even imagine health care without plastics, and I don't even think we should go there," Dr. Jodi Sherman, founding director of the Yale Program on Healthcare Environmental Sustainability, told Live Science. "I would argue that plastic has allowed very important innovation of medical devices and supplies, and is here to stay."

Right now, primary plastics are "ridiculously and unsustainably cheap," so oil-free alternatives can't compete on cost, Bauer said.

Bioplastics, made from crops, could provide a way forward, MacDonald said. But the story of biofuels serves as a cautionary tale. Soybean fields have taken over large swaths of U.S. farmland — in part because of its use as a biofuel.

"We have a finite amount of agricultural land," MacDonald said. "If we turn a lot of it over to growing fuels, what do we do about feeding people? It's not an easy equation. Everything is related and interlinked."

The beginning of the end of oil

"The oil industry isn't going to collapse because we run out of oil, there's plenty of oil left," MacDonald said.

But at some point, clean energy technologies will become so cheap that it won't pay off to drill and extract oil.

The first method to be phased out will be wildcat drilling, in which an area with unproven reserves is explored, MacDonald said. This is risky and extremely costly if you don't find anything. Even drilling new wells in areas with known oil reserves is eye-wateringly expensive: Companies spend tens to hundreds of millions to get wells and rigs staffed up and running, and then it's years before they turn a profit.

"You're spending money like a drunken sailor in the hope you're going to get some money back," MacDonald said. "It ain't a quick process. That's why oil companies are big — they have to be as they're carrying a huge amount of risk."

Still, oil wells will continue to pump in the vast sand fields of Saudi Arabia for decades. In the U.S., production will continue at high levels through 2050.

Femke Nijsse, a complexity scientist at the University of Exeter in the U.K. whose research focuses on modeling climate, energy systems and the economy, told Live Science she's hopeful that global oil use will be cut by 95% by 2065, with aviation and shipping as the remaining strongholds.

MacDonald predicts a "less spectacular" decline, falling a quarter by 2050. "At some point you'll get to a cliff where it will go down quite rapidly," he added.

Some experts can't imagine a post-oil future at all. Kevin Book, managing director of ClearView, a research firm that looks at energy trends, told Live Science that artificial intelligence and geoengineering will change oil extraction and refining, but that oil won't disappear until a technology that doesn't exist yet, like fusion energy, makes it obsolete.

But the push for decarbonization means oil will eventually become a flash in the pan in our history. Like industrial whaling, our taste for it will dissipate until just a few small strongholds remain.

Fifty to 100 years from now, oil derricks and drilling fields in the U.S. may start to look like the abandoned mine museums and gold-rush ghost towns that litter the American West — tourist attractions that paint a picture of a lost way of life, an economy firmly in the past.

When the Spanish first reached the Andes, they found something surprising: Many of the locals had long, pointy heads. They discovered that the Collagua, an indigenous group in Peru that was conquered by the Inca, had a practice of shaping the head starting in infancy, before the skull bones fused and soft spots disappeared.

The Spanish jumped to the worst conclusions.

"They said it was this horrible thing and brains bled out of ears," Christina Torres, a bioarchaeologist at the University of California, Riverside, told Live Science. "But that doesn't seem to be the case."

Prehispanic Indigenous groups were not the only ones to practice head shaping. For centuries, archaeologists have found skulls on every continent except Antarctica that show evidence of "cranial vault modification" — heads shaped to be either flatter or more conical than they would be if left alone.

Given that babies cannot bind their own heads, experts think head shaping was done by caregivers. Now, archaeologists are beginning to uncover clues about why people performed this practice for millennia, particularly in places like the Andes, where the practice has been documented the best.

Through systematic analysis, what experts are uncovering is a profusion of practices and explanations, some of which are baffling or contradictory. In some places, a shaped head may be a marker of group status, while in other places, head shapes differ even among close family members. And in other places, the feature used to identify it — the unusual head shape — may not even have been the intent of the practice, researchers are finding.

"Something as ostensibly shocking as cranial modification may have been almost a routine practice for some children in some time periods," Matthew Velasco, a bioarchaeologist at the University of North Carolina at Chapel Hill who studies head shaping in the Peruvian Andes, told Live Science.

What's more, it likely originated very deep in human history — and emerged in many times and places, Velasco said. "I think we have to start from the assumption that the meaning varies across time and space."

How are heads shaped?

Bone remodels easily when children are young, so a simple strip of wrapped cloth can control how the head grows, much like how a bonsai tree can be shaped and pruned, Torres said. For example, nowadays, babies with plagiocephaly — a flat spot caused by sleeping in one position — are often prescribed helmet therapy to change their head shape.

Experts have identified more than two dozen apparatuses that were used to create different head shapes, but "the most typical method would be just wrapping the baby's head circumferentially and making a longer, more conical shape," Torres explained, as this technique requires the least equipment and the least training.

Based on historical records from groups that practiced it, head wrapping began by around 6 months of age in most cultures and ended within a year or two, Tyler O'Brien, a forensic anthropologist at the University of Northern Iowa, wrote in "Boards and Cords" (Bloomsbury Publishing, 2024), a book about the worldwide history of cranial modification. The shaping was likely done by a mother or midwife.

There's not a lot of information, though, on whether this practice was painful, but it does not appear to have had any major consequences for brain development, Torres said.

"There's one instance where [bioarchaeologists] think a child died because of cranial modification, where the head was compressed too much," she said. "That is anomalous, as far as I can tell. This was a slow and gradual process done with fabric and pillows."

So, although Spanish explorers in the Andes said they were shocked by "brains coming out" and explorers in Borneo and Vanuatu said kids' eyes "bulged from their sockets," these subjective accounts are probably greatly exaggerated, O'Brien wrote. In reality, the child likely adapted quickly to any discomfort, and the brain would have conformed to the shape of the skull, resulting in no ill effects on cognition or intelligence.

If not done properly, however, head shaping that involved overly restrictive or infrequently changed bindings could cause infection. "I think the worst thing you could have is a [skin] ulcer that gets infected and then eats through the bone, which does happen," Christine Lee, a bioarchaeologist at the University of Mississippi, told Live Science. Scalp infections and temporomandibular joint (TMJ) issues also may have occurred, Torres said.

Archaeological evidence for head shaping

To identify a shaped head, archaeologists sometimes used visual inspection as well as human skull measurements, known as craniometry. Craniometry has been around since the 19th century, when American naturalist and eugenicist Samuel Morton used cranial traits to create racial hierarchies that have since been debunked.

Although there is no standardized, agreed-upon method to determine if a head has been shaped, archaeologists often use a 3D, mathematical analysis of cranial measurements to see whether the ratio of certain measurements, such as the width, length and height of the skull, are outside the range of what is expected as part of natural variation and are thus likely to have been intentionally shaped.

That analysis suggests that head shaping is widespread in the archaeological record. Cranial modification has been found in skulls from Europe, the Near East, Africa, Asia and Oceania, and it is most strongly associated with the Americas. But that doesn't necessarily mean it was more common there; rather, evidence of shaped heads may have been better preserved in the Andes, where the cool, dry conditions did not degrade mummified remains as quickly, Velasco said.

In fact, the oldest archaeological evidence of head shaping comes from Australia. Two artificially flattened skulls were discovered in the southern state of Victoria at the site of Kow Swamp, which is at least 13,000 years old.

And ancient skulls reveal the practice boomed in the Neolithic period, appearing in Europe around 12,500 years ago, in China around 11,000 years ago, and in what is now Iran around 10,000 years ago, according to O'Brien.

Early accounts

Some of the earliest explanations for head shaping, which come largely from Spanish explorers in the Americas, are some of the most suspect. That's because few explorers bothered to ask practitioners why they engaged in head shaping. Instead, these wild stories were often based on rumor or hearsay.

For instance, Christopher Columbus first reported head shaping among the Indigenous people of Hispaniola, the island that encompasses the Dominican Republic and Haiti, in 1492. He illogically guessed that the islanders had flat heads because their mothers pressed them tightly between two wooden planks, causing the skull bones to thicken like helmets and protect them from Spanish blows, according to Pilar Zabala Aguirre, an anthropologist at the Autonomous University of Yucatán in Mexico, who has compiled more than 100 Spanish historical records on the practice.

Other explorers invented different possible explanations: ethnic grouping; high military rank; attributes such as courage, bravery or obedience; the ability to carry heavier loads strapped around the forehead; health improvements; and beauty ideals, Zabala found.

These explanations are even more suspect because they were often tied to racism or beliefs in the superiority of Western culture or even used to explicitly make that argument.

For instance, English physician John Bulwer cataloged various types of body modification in his 1650 book "Anthropometamorphosis," condemning them as disfiguring and an affront to God, according to O'Brien.

It wasn't until the early 20th century that anthropologists moved away from "studying abnormal head shape in the living 'other' and describing it as hideous, frightful, and disgusting," O'Brien wrote, and toward a less-biased understanding of cranial variation.

Changing understanding of the practice

Using those robust, less biased methods, archaeologists are gradually unwrapping some of the mystery surrounding the practice, mostly in the Americas. And what they're finding is not an overarching trend, but a range of reasons and practices.

For instance, the Collagua in Peru ostensibly "told the Spanish that they shaped the heads of their children like the mountain from which they come," Velasco said.

Among the Indigenous Caddo people of Oklahoma, meanwhile, different kinds of shaping reflected membership in different clans, Lee said.

Yet head shaping differed not only within cultures but even within families. Velasco's ongoing research, which involves analyzing the DNA of extended families buried together in the Andes, has revealed that the heads of biologically linked people were often shaped in different ways — so one family member might have an elongated head, while another might have an unmodified, rounded head.

In fact, in some cultures or families, the shapes of the heads may have been the unintended result, rather than the goal, of a practice that was more important to them, such as binding.

"The shape itself might actually be collateral to the practice" in the Andes, Torres said.

Instead, in some parts of the Americas, head shaping may have lingered simply as a tradition. For instance, in the Andes, the practice may be part of a rite of passage for either the infant or the mother, Torres said. If head shaping began around 6 months of age, that is a time when the baby's first teeth were coming in and weaning foods may have been introduced. There, head shaping may be similar to putting jewelry on a child or baptizing them to protect them, she said.

Among the pre-hispanic people of the Andes, "it's basically a child-rearing practice," Torres said. "In the same way that some people swaddle their children, the same way that there's religious circumcision, you bind the heads of your children because that is what we do to our children."

A natural conclusion

In fact, the idea of shaping a head into a conical form may have been presented by birth itself. The infant cranium naturally deforms when it passes through the birth canal, Velasco said.

"When my child was born, for example, he had a slightly conical head," he said. "Birthing presents this possibility to every parent, and it doesn't take much of a leap to feel the supple head of a child and to wrap it, to clothe it."

Thus, it's not surprising that many cultures might have stumbled upon head shaping, given that "anyone who has observed or assisted a human birth will recognize that the human head is malleable," Velasco added.

This recognition of the plasticity of a baby's head may have spurred a need to protect it. For example, among the Maya, modifying an infant's head was likened to putting a roof on a house and was thought to protect the child.

That protection may have then become more metaphorical than physical. In some places, it has "an almost talismanic aspect to it," Lee said, as if something bad might happen to the child if the shaping were not done. "That implies there must be almost a fear of not doing it."

Beauty and 'in-group' status

Outside the Americas, few historical records describe head shaping, but "it seems to have been independently invented in multiple places," Torres said.

Each of these places may have had different justifications for the practice. In prehistoric China and Japan, for instance, head shaping was likely a status marker tied to the elite. Lee thinks it's likely that in ancient Asia, head shaping — similar to foot binding in more recent times — represented an extreme way of achieving a beauty standard.

Similarly, during the fourth to seventh centuries in Europe, head shaping surged in popularity among the Huns, skeletons reveal. Without historical evidence from the Huns detailing the reason for this practice, experts have conjectured that it was a "fashion wave" in the Eurasian steppe that conferred higher social status.

But head shaping often leaves subtle traces, meaning people who engaged in the practice may not have looked much different from those who didn't. That suggests the practice was not necessarily a striking visual marker of in-group status, which raises questions about whether that was its purpose.

"Hair hides a lot," Lee said. "There are people today with unusually shaped skulls, and it just doesn't show depending on their haircut." For example, venture capitalist Marc Andreessen and political consultant Roger Stone both have (presumably naturally) pointier-than-average heads.

Modern cases of head shaping

Although cranial vault modification has been going on for tens of thousands of years, it persisted well into the 20th century in parts of Africa, Oceania and Europe.

The Arawe people in Papua New Guinea practiced head shaping as late as the 1930s, according to a study of several villages in 1955. The Arawe used bark-cloth bandages to create what they considered an aesthetically pleasing "long-headed" style.

Among the Mangbetu people of the Congo, the practice of "lipombo" involved tightly wrapping an infant's head with cloth bandages to encourage a long, conical shape thought to be beautiful and powerful. The Belgian colonial government outlawed the practice, which died out in the 1950s.

And in early 20th-century France, some parents chose to band their newborns' heads immediately after birth for up to four years in a practice called "bandeau," which practitioners thought protected babies from injury. The "Toulouse deformity," named after the region where it was practiced, has been captured in a series of historical photos, but bandeau declined in popularity and disappeared by World War I.

"This is not something that is simply a brute barbarian practice that people evolved out of," Velasco said. "It is fairly independent of social complexity."

The practice of shaping an infant's head to look like an "alien's" may seem foreign or bizarre to us in the 21st century. But the human body has long been a canvas for cultural, spiritual and personal expression.

The earliest known tattoos date back at least 5,000 years in Copper Age Europe, tooth filing and "grills" go back 2,000 years to the Maya, and neck elongation was practiced 1,000 years ago in Southeast Asia. Today, we tend to modify the soft tissues of our bodies through common practices like ear piercing and circumcision, but also more uncommon procedures like horn implants, eyeball tattooing and Brazilian butt lifts.

"Cranial modification is part of a practice that is universal: body modification and presentation," Velasco said. "We all invest our future in our children in different ways. That's how I think about cranial modification. It's like an investment in the future of a child. And when you put it in that way, it's hard not to relate to it."

For researcher Lynnette Averill, the quest to find a treatment for post-traumatic stress disorder (PTSD) is deeply personal. Averill's father served as an enlisted infantryman with the U.S. Marine Corps in Vietnam and struggled to cope with his war experiences when he returned home. After years of ineffective treatments, he died by suicide when Averill was three.

Driven by a mission to support veterans' mental health, Averill trained as a psychologist and began working with people with PTSD — a condition that affects more than 12 million Americans in any given year. Victims of violence, abuse and accidents can experience post-traumatic symptoms such as persistent flashbacks, hypervigilance, and entrenched negative beliefs about themselves and their environment.

"People can be very stuck in black-and-white thinking, such as, 'I'm a bad person,' 'I deserve this,' 'the world is dangerous," Averill, a clinical research psychologist at the Baylor College of Medicine in Texas, said during a panel discussion at the Psychedelic Science conference in Denver in June 2025.

The root of these symptoms lie in how trauma shapes changes in the brain in the weeks and months after a frightening event. The brain's fear center — the amygdala — becomes hyperactive, constantly signaling danger, while the brain regions responsible for contextualizing memories and managing emotional responses become less active and less able to counterbalance those fear signals. Traditional therapies, such as antidepressant medications and trauma-focused psychotherapies, help only a fraction of patients and can take months to be effective.

"For many people with PTSD, they simply aren't enough, " Averill told Live Science.

Consequently, Averill is one of a group of researchers who are exploring a new potential avenue for treating PTSD: psychedelics. Psychedelic-assisted psychotherapy, using MDMA or psilocybin, may act on the brain systems disrupted in PTSD, rather than simply treating the symptoms.

The early findings have been positive: A recent clinical trial showed that 67% of patients who received MDMA-assisted therapy no longer met PTSD criteria after treatment, compared with 32% in the placebo group, and clinical trials investigating psilocybin's potential to treat the condition are showing promise.

Averill is currently leading a pioneering Texas state-funded clinical trial investigating psilocybin for veterans with PTSD and has seen how quickly the drugs can act.

"There's potential for people to feel that the needle has moved in hours," Averill said. "And that is just quite literally lifesaving."

How trauma changes the brain

PTSD shares symptoms with depression and anxiety. Yet it is characterized by a response to a single trauma or set of traumatic events. Such experiences spark fear and often challenge an individual's core beliefs that the world is a just, safe and predictable place. People with PTSD can feel helpless and without agency.

It's normal for people who have endured traumatic events to experience these symptoms for a short time, and for most people, they resolve within a week or two, clinical psychologist Gregory Fonzo, a co-director of the Charmaine and Gordon McGill Center for Psychedelic Research and Therapy at the University of Texas at Austin's Dell Medical School, told Live Science. But a subset of people get stuck.

"PTSD is essentially a disorder of nonrecovery," Fonzo said.

During the instigating event, trauma activates the brain's fear alarm system, including the amygdala, and signals a rapid release of stress hormones and neurotransmitters, such as norepinephrine (noradrenaline). The higher levels of norepinephrine increase arousal and the physiological “fight or flight” response. The interaction of norepinephrine and corticotropin-releasing hormone modulates increases activity between the amygdala and hippocampus, reinforcing the synaptic connections that help store fear and vivid details of the event.

For the majority of people who recover quickly, this powerful fear response is successfully extinguished. The prefrontal cortex, the brain's regulatory control system, and the hippocampus, the brain's center for memory, exert top-down control and file the event as a past memory and no longer dangerous, which restores normal signaling in the amygdala.

For people who develop PTSD, however, the trauma creates more persistent changes in the brain, Dr. Jerrold Rosenbaum, a psychiatrist and director of the Center for Neuroscience of Psychedelics at Massachusetts General Hospital, told Live Science.

In PTSD, the amygdala remains stuck in an overactive state, causing symptoms like hyperarousal, irritability and being easily startled. At the same time, the prefrontal cortex, which normally calms those alarms, becomes underactive, leaving the amygdala's overreactive fear response unchecked.

Neuroimaging has shown that PTSD is associated with a reduced volume of the hippocampus, which is the brain region that processes the context — the where, when and circumstances — of an event. In normal circumstances the hippocampus can discriminate between real and perceived danger. For example, it will categorize the sound of a car backfiring in an everyday environment differently than the blast of gunshot, which occurred specifically in a war zone. But a diminished hippocampus could make it harder for patients to distinguish between the two.

There's also evidence that PTSD is associated with changes in the connectivity of the default mode network (DMN), a set of interconnected brain regions that are highly active when a person is resting, their mind is wandering, or they are engaged in thinking about themselves or their memories rather than the outside world. Researchers hypothesize that heightened connectivity within the DMN is a neurobiological abnormality that pushes the system into overdrive, potentially leading to rumination or the involuntary re-experiencing of negative events (or flashbacks), which are characteristic symptoms of PTSD. In PTSD, the DMN also appears to be abnormally disconnected from the executive control regions and simultaneously abnormally connected with lower-level emotional systems, such as the amygdala. Scientists believe this increased coupling may drive intense, automatic emotions such as shame and guilt.

In a healthy brain, the prefrontal cortex can evaluate and change upsetting thoughts, but PTSD compromises this ability. PTSD is also associated with reduced levels of brain-derived neurotrophic factor (BDNF), a signaling protein that is critical for neural plasticity — the formation of new synapses and the strengthening or pruning of existing neural connections. The deficit in BDNF effectively "locks" the trauma response in place and can prevent the brain from integrating new, non-fearful thinking, scientists theorize.

The result is a system trapped in fearful thinking and the past.

"PTSD can definitely manifest in rigidity and entrenchment of ways of thinking about oneself and others in the world," Brandon Weiss, a psychologist at the Johns Hopkins Medical Center for Psychedelic and Consciousness Research, told Live Science.

Treatment challenges

Physicians have used trauma-focused psychotherapy, such as cognitive processing therapy (CPT) and prolonged exposure (PE), to treat PTSD. In CPT, therapists guide patients to examine and change distorted thinking, such as the belief that they are to blame for events outside of their control. PE focuses on teaching the patient to reduce their fear by gradually confronting the traumatic memories in a controlled environment.

But these treatments are difficult for people to go through because they have to face the stuff that is really bothering them, Fonzo said. "So a lot of people drop out before they finish the treatment."

Access to specialized, trauma-focused therapists can be challenging, and not everyone benefits from such treatments, Fonzo said. Other treatment options include SSRIs, like sertraline (Zoloft) or paroxetine (Paxel), but only 20% to 30% of patients experience a remission on these drugs.

Most importantly, many patients feel like the medication hasn't addressed the root cause of their trauma, Rosenbaum said.

Creating a safe space to process trauma

For Jennifer Mitchell, a neuroscientist and the associate chief of staff for research and development at San Francisco Veterans Affairs Medical Center, it was imperative to find effective treatments for military veterans suffering from PTSD.

"I see how debilitating it is for individuals that have served and have experienced combat," Mitchell told Live Science. "And we absolutely owe them to come up with better treatments."

Nearly a decade ago, Mitchell and her colleagues began to investigate whether MDMA (3,4-methylenedioxymethamphetamine), commonly known as ecstasy or molly, could treat PTSD when used in conjunction with psychotherapy. In 2023, the team published their findings on a diverse group of 104 participants with PTSD from the general population, 80% of whom had a history of contemplating suicide. By the end of the study, 71% of the participants who'd received MDMA no longer met the criteria for PTSD, and another 15% still had symptoms but had what the researchers termed a "clinically meaningful benefit."

MDMA's effectiveness hinges partly on its ability to act as a neuroplastogen — a compound that harnesses the brain's ability to form new connections and strengthen and reorganize existing connections. In the past few years, scientists have begun to explore the potential for another neuroplastogen to treat PTSD: psilocybin.

Clinical trials have found psilocybin, the main psychoactive ingredient in magic mushrooms, to be a promising therapy for treatment-resistant anxiety and depression, and studies conducted on animals have pointed to its potential for treating PTSD.

A study on mature adult mice demonstrated that MDMA temporarily reopens a critical period for where the brain is sensitive to learning that social behaviors are beneficial by inducing structural and functional changes in the brain’s reward circuits. Specifically the drug makes the nucleus accumbens reward circuitry more sensitive to the social hormone oxytocin. This enhanced sensitivity allows the adult brain to re-encode social cues as intrinsically rewarding and safe, facilitating the re-learning of trust and attachment for up to two weeks after a single dose, the researcher theorize.

Mitchell has witnessed similar responses in humans and hypothesizes that the drug creates a neurobiological state in which patients can form a strong, trusting bond with a therapist. "There's a therapeutic window where people feel renewed energy, they don't feel so stuck, and they can actually work on the psychological side of their issues," Mitchell said.

Simultaneously, functional neuroimaging data points to MDMA's impacts on the fear circuitry of the brain. The drug decreases activity in the amygdala while increasing activity in the prefrontal cortex. MDMA also restores normal levels of BDNF in the amygdala, hippocampus and prefrontal cortex, theoretically leading to the formation of new synapses and structural changes that enable greater plasticity in these regions.

Researchers hypothesize that the dampening of the fear response, the enhancement of the prefrontal cortex's regulatory role, and the restoration of flexibility in key brain regions set the scene for patients to revisit and reprocess memories without being overwhelmed by fear. By entering therapy with this more regulated perspective, recovery can happen quickly.

"By the end of a treatment session, you can see that something has shifted. The subject will be holding themselves differently and look hopeful," Mitchell said. "I've been doing this type of research for 35 years, and it is the most remarkable thing that I've seen."

Once the patients have analyzed and processed their trauma, they are less likely to slip back into their PTSD symptoms after MDMA, Mitchell said. The researchers collected data on the patients in increments over the two years after their treatment ended, and the positive benefits appear to be durable, Mitchell said.

Breaking free

When a single dose of psilocybin was injected into chronically stressed mice that had developed learned helplessness and avoidance behaviors, for example, researchers could see immediate and enduring structural changes in the rodents' brains. The rapid increase in dendritic spines — the tiny protrusions that form synapses, the connections between brain cells —suggests that psilocybin may directly reverse the loss of neuronal connections observed in the frontal cortex of rodents subjected to chronic stress, the authors suggest. And the changes could prepare the brain for fear extinction and emotional processing —key elements of overcoming trauma.

Consequently, several trials are underway to investigate psilocybin as a treatment option for PTSD. In August 2025, biotechnology company Compass Pathways published its findings on a trial designed to test the safety of psilocybin for PTSD. The small safety study wasn't designed to measure effectiveness. Nevertheless, participants seemed to show an immediate reduction in PTSD symptoms after a single 25-milligram dose of the company's synthetic psilocybin. Clinicians reported the improvements had endured when tested 12 weeks later.

In the study Averill is leading, clinicians started dosing seven participants who had experienced PTSD symptoms for an average of 19 years at the start of the trial in February. The early findings "have been incredible," Averill said during a panel discussion at the Psychedelic Science conference in Denver in June 2025.

Within a few hours of the first dose of psilocybin, every single veteran reported positive changes in their beliefs and perceptions. After the treatment effects subsided, participants underwent therapy and said the drug enabled them to reevaluate their original traumatic experiences from a nonjudgmental perspective, without the shame and guilt, Averill said.

So how does psilocybin bring about these changes?

Studies in mice have demonstrated that, in a similar vein to MDMA, psilocybin helps brain cells grow new dendritic spines. These new branches appear in the prefrontal cortex and the hippocampus, the key regions responsible for learning, planning and memory.

But brain scans from people have shown that psilocybin simultaneously disrupts and desynchronizes, or dissolves, connectivity within the DMN for three weeks. Scientists hypothesize that this desynchronization of the brain system involved in self-referential processing results in a mind that is less constrained, more flexible and less consumed by self-reproach. They propose that the disruption of rigid cognitive patterns and negative thought loops may promote psychological flexibility and self-compassion.

"It has been surprising to see these maladaptive beliefs shift when I'd learned that change would only come about through longer term talk therapy," Weiss said.

Dissolving rigid thinking can open up the mind to new possibilities, including the idea that the patient wasn't responsible for the original trauma, Weiss said. Weiss is currently carrying out a clinical trial examining the effect of psilocybin therapy on the cognitive beliefs related to PTSD and is witnessing participants' experiences of the treatment.

"Many are endorsing less guilt and are able to let go of the sense that they were to blame for this happening," he said.

Proceeding cautiously, with haste

When Fonzo first reviewed the data on psychedelics as a treatment option in mental health, he was excited by what he found. But he pointed out that there haven't yet been any large, controlled studies evaluating psilocybin for PTSD. The research is still in its early stages with small pilot studies or clinical trials still assessing the safety of the drug. "There needs to be a sober perspective on what the evidence does, and does not show," Fonzo said, "because these treatments aren't necessarily going to be suitable for everyone."

Fonzo believes the answer lies in the expansion of funding for clinical trials, but research in psychedelics still faces steep hurdles. Both MDMA and psilocybin are listed as Schedule I substances in the U.S. — a federal classification reserved for drugs considered to have a high potential for abuse and no accepted medical use. That label makes studying them a bureaucratic nightmare, as researchers must navigate complex regulatory approvals and secure special licenses just to handle the compounds.

On top of that, every trial demands careful screening to find participants who meet the strict inclusion and exclusion criteria. Also, due to the profound and unmistakable psychoactive effects of psychedelics, in trials, it can be hard to hide who is and isn't getting the drug and thus create a robust placebo group.

Despite these challenges, new clinical trials are underway to explore variations in dosage, psychotherapy pairing and long-term outcomes. Weiss and his colleagues are investigating administering combinations of MDMA and psilocybin — so-called psychedelic stacking — and comparing the effectiveness of each treatment separately. "We're still learning about what is the most efficient and rapid-acting approach to helping people," Weiss said.

For veterans who are experiencing suicidal thoughts as part of their PTSD, finding interventions that spur rapid change could be key, Weiss said. Despite the compelling findings, the regulatory approval of psychedelic treatments for PTSD is moving slowly. In August 2024, the U.S. Food and Drug Administration declined to approve MDMA-assisted therapy for PTSD, citing concerns over the study design and blinding procedures. Mitchell has been frustrated about the decision not to approve the treatment with guardrails, or for a subset of people who really need it.

"My own brain gets cranky when I hear people not consider PTSD a life-threatening condition — because it is," Mitchell said. In the United States in 2024, an average of 17 veterans died each day by suicide.

"That's why speed matters. We can't wait months for treatments that barely work," Mitchell said.

Averill has witnessed PTSD patients liberated by psychedelic therapy, but she cautioned that the psychedelic experience is not a "silver bullet" but rather one that opens the doors to healing. One veteran in the clinical trial she'd been leading described feeling like he'd been locked in a cage and said every movement felt painful. The psilocybin removed the cage, and he was able to revisit his traumatic experiences.

Averill's goal is to continue exploring how psychedelic-assisted therapy can help PTSD sufferers move beyond merely tolerating their existence. "We want to help people move forward and build lives they really want to be living," Averill said.

Colossal monsters lurk in the centers of all galaxies. Known as supermassive black holes, these gravitational beasts can have millions to billions of times more mass than the sun.

For decades, astronomers have wondered where these behemoths came from and how they got so huge. Early on, physicists thought that supermassive black holes formed like other, smaller black holes do — with large stars collapsing and becoming sun-size black holes that slowly devoured surrounding matter and merged with one another over billions of years.

But it has become increasingly clear this model is broken.

The James Webb Space Telescope (JWST) has peered back to some of the earliest epochs in cosmic history to spot gigantic black holes that are too big, too early to be explained by traditional models. Researchers are beginning to piece together a story of how they originated, which likely involved strange and exotic processes.

Emerging research suggests enormous black holes could have existed since the universe's earliest days, perhaps even before stars and galaxies, and that they came about in multiple ways. While future discoveries will help narrow down the predominance of each formation mechanism, many in the field are already thrilled to be chipping away at a long-standing cosmic mystery.

"This is one of the most exciting phases of my career," Roberto Maiolino, an astrophysicist at the University of Cambridge, told Live Science. "I'm tempted to call it a real revolution in our understanding of the formation of these objects."

Mystery giants

Hints of the cosmic size discrepancy arose in the early 2000s, when instruments like the Sloan Digital Sky Survey helped capture data on tens of thousands of extremely bright objects called quasars in the faroff universe. These luminous entities are thought to be gargantuan black holes in the centers of galaxies. They feed on vast amounts of gas and dust, and then spew powerful radiation. The Sloan survey showed that many quasars existed when the universe was just 800 million years old — a fraction of its current 13.8 billion-year age. The existence of these behemoths, which have millions to billions of times the sun's mass, was a head-scratcher for cosmologists.

That's because a typical black hole arises when a huge star nears the end of its life and explodes as a fiery supernova. The core of the titanic star collapses into a superdense point from which nothing, including light, can escape. Such stellar-size black holes are generally around 10 to 100 times as massive as the sun. While these objects can become gravitationally attracted to one another and merge into ever larger black holes, there didn't appear to be enough time for such processes to build them up into quasar-scale territory at the earliest points in cosmic history.

"We knew that either they grow very fast or there had to be some other ways of forming them," astrophysicist Ignas Juodzbalis, also of the University of Cambridge, told Live Science.

The question was how. One leading theory posits that, in the past, ginormous clumps of gas and dust could collapse under their own weight, rapidly forming a black hole with perhaps 1,000 to 1 million times the sun's mass. These direct-collapse black holes, as they're called, would then grow by feeding on gas and dust and merging into the supermassive black holes seen in today's galactic centers.

Models predicted that as such black holes gorged, they would become extremely bright compared with their host galaxies, either matching or topping surrounding stars' luminosities. In other words, they would become quasars.

In 2023, JWST spotted a distant galaxy, dubbed UHZ1, that seemed to align neatly with the direct-collapse black hole model. The galaxy existed when the universe was a mere 470 million years old and contains a black hole with an estimated mass of 40 million suns.

Astronomers lucked out because UHZ1 was spotted both by JWST, which sees in the infrared part of the electromagnetic spectrum, and by NASA's Chandra X-ray Observatory, which sees in X-ray light. Infrared light mostly comes from stars and warm dust heated by starlight, whereas the more powerful X-rays blast out from the devouring black hole.

And UHZ1's infrared and X-ray brightness are quite similar to one another, which suggests a black hole so large that it rivals the mass of all the stars in its galaxy. (For comparison, a modern galaxy like our Milky Way has around 20,000 times more mass in its stars, gas and dust than in its central black hole.) No one had ever seen anything like this before.

But researchers had predicted exactly how the colors emitted by a direct-collapse black hole would appear in JWST's instruments, along with several other key properties that could identify such an object .

"It turns out that UHZ1 remarkably satisfies all these properties," Priyamvada Natarajan, an astrophysicist at Yale University and lead author of the paper making those predictions, told Live Science.

Little red dots

UHZ1 is not alone. From almost the moment it turned on, JWST has been detecting extremely compact red entities that existed mainly when the cosmos was between half a billion and 1.5 billion years old. Known as "little red dots," they were originally thought to be galaxies far too big to have formed in the early universe, leading some scientists to call them "universe breakers" for upending models of cosmic history. The prevailing consensus is now moving toward the possibility that, rather than unusually large galaxies, these are bizarre, humongous black holes.

For instance, an object called QSO1 that existed when the universe was around 700 million years old has been studied intensely since it was discovered in 2023. A recent investigation looked at gas swirling around QSO1's center to try to pin down its mass with high precision. Swirling gas travels at a certain speed depending on the gravitational force tugging it as it spins. Using this technique, astronomers have shown that QSO1's mass is around that of 50 million suns. Moreover, all of the mass appears to be in a compact region around the black hole, with very little evidence of a large stellar population.

"We still don't see where the host galaxy is," Lukas Furtak, an astronomer at the University of Texas at Austin, told Live Science. "There doesn't really seem to be one."

This prospect — a gigantic black hole with no visible host galaxy — has been conjectured but never previously observed. Yet that appears to be what many of these little red dots are. Another recent study analyzed an object named "The Cliff," which likely weighs billions of times as much as the sun and is from about 1.8 billion years after the Big Bang. JWST's data showed a very sharp jump in The Cliff's light at a narrow wavelength that usually arises from dense hydrogen gas at a specific temperature. The findings indicate that The Cliff might be a long-hypothesized object called a quasi-star or a black hole star.

A quasi-star would be a potential stage in the evolution of a direct-collapse black hole. After the central huge chunk of gas crumpled to form a black hole, an outer sphere of gas and dust would remain, get heated by the black hole's emissions and glow in red wavelengths. This entity would look somewhat like a giant red star but would in fact be an envelope of hot hydrogen gas cocooned around a supermassive black hole.

In the very beginning

While direct-collapse models can explain a lot of what JWST is seeing, there remain a few other possibilities for supermassive black hole formation.

First proposed by Stephen Hawking in the 1970s, primordial black holes are a class of objects that could have arisen in the first few moments after the Big Bang, when dense regions collapsed under their own weight. Such black holes could come in a wide range of sizes, including ones large enough to act as the initial seeds for later supermassive black holes. One study has shown that mergers of primordial black holes could explain GN-z11, a galaxy from when the universe was a mere 400 million years old that contains a black hole with an estimated mass of 2 million suns.

Another theory has posited the existence of "not-quite-primordial black holes ." These would have come about within the first few million years after the Big Bang — later than primordial black holes but still long before any stars — when large clouds of hydrogen and helium collapsed under their own weight.

"For primordial black holes, you need these really extremely dense regions in the very early universe," Wenzer Qin, a theoretical physicist at New York University, told Live Science. That generally requires a lot of fine-tuning of parameters in a cosmological model, she added. When you relax such tight constraints a bit, dense regions appear at slightly later times in cosmic history, creating direct-collapse black holes that can go on to merge and end up as supermassive black holes.

Astronomers think that almost all elements heavier than hydrogen and helium were created in the nuclear bellies of giant stars and were then strewn about the universe when those stars went supernova. Many of the early black holes and young galaxies that JWST is seeing contain low amounts of these heavy elements. That could suggest that at least some of these objects formed from either primordial or not-quite-primordial black holes, given that both would have arisen long before any stars existed.

Researchers are still debating which of these models might be dominant for monster-black-hole formation, but most favor a blended view.

"I think, in the end, it will be some combination of all these mechanisms that gives rise to the entire population of supermassive black holes," Qin said.

Other missions such as the European Space Agency's Euclid observatory, launched in 2023, and NASA's Nancy Grace Roman Space Telescope, expected to launch in 2027, will team up with JWST to discover and study more early supermassive black holes. That should help researchers differentiate between these formation mechanisms and determine which, if any, is more common.

One thing that appears to be growing clearer to many astronomers is that supermassive black holes in the centers of galaxies probably didn't come from stellar-size ones.

Thanks to its unparalleled abilities, JWST has upended our understanding of early cosmic history and is helping to rewrite the story of how gigantic black holes may have developed.

"The universe is littered with supermassive black holes that form extremely early," Natarajan said. "I can't tell you how exciting that is."

The group had traveled for thousands of miles, crossing Africa and the Middle East until finally reaching the dimly lit forests of the new continent. They were long-vanished members of our modern human tribe, and among the first Homo sapiens to enter Europe.

There, these people would likely have encountered their distant cousins: Neanderthals.

These small bands of modern-human relatives had hooded brows, large heads and squat bodies, and they had spent epochs acclimating to Europe's colder climate. At several points across millennia, these two forms of humanity would meet, mingle and mate.

Tens of thousands of years later, these ancient encounters have left traces in the genetic code of billions of humans alive today. The lingering genes affect us in ways large and small, from our appearance to our risk of disease.

"In some places in our genome, we're more Neanderthal than we are human," Joshua Akey, a professor of integrative genomics at Princeton University, told Live Science.

These were our closest human relatives, and this is their legacy.

The first encounter

By 75,000 years ago, but possibly up to 250,000 years ago, the ancestors of most modern Eurasians first ventured out of Africa and into Eurasia. Here, modern humans came face-to-face with Neanderthals, who last shared a common ancestor with modern humans hundreds of thousands of years earlier and had been living in these continents ever since. On multiple occasions over the millennia, the groups interbred.

Related: Could Neanderthals talk?

At first, modern humans inherited whole chromosomes from Neanderthals, Sriram Sankararaman, a professor of computer science, human genetics and computational medicine at UCLA, told Live Science. However, from generation to generation, via a process known as genetic recombination, these stretches of DNA were broken up and shuffled around.

Neanderthal DNA was generally "deleterious" to modern humans, meaning it was rapidly weeded out of modern humans' DNA through evolution. This resulted in "deserts of Neanderthal DNA," or large regions of the modern human genome lacking it, Sankararaman said. For instance, scientists think the Y chromosome in males doesn't contain any Neanderthal genes. It may be that genes on the Neanderthal Y were incompatible with other human genes or they may have been randomly lost via a process known as genetic drift.

In people who inherited Neanderthal DNA, the X-chromosome also contains a lot less Neanderthal ancestry than other, non-sex chromosomes carry. This is probably because any harmful or nonfunctional mutations on the X chromosome will be expressed in males, because they lack a matching, functional copy of the gene to compensate. That likely created strong evolutionary pressure to remove such harmful Neanderthal genes from the modern human X, Emilia Huerta-Sanchez, an associate professor of ecology, evolution, and organismal biology at Brown University, told Live Science.

But some Neanderthal DNA helped modern humans survive and reproduce, and thus it has lingered in our genomes. Nowadays, Neanderthal DNA occupies, on average, 2% of the genomes of people outside Africa. However, the frequency of Neanderthal DNA that codes for beneficial traits may be as high as 80% in some regions of the genome, Akey said.

Our physical appearance

For many people, the legacy of Neanderthals is apparent in a highly visible feature: skin color.

A Neanderthal gene variant on chromosome 9 that influences skin color is carried by 70% of Europeans today. Another Neanderthal gene variant, found in most East Asians, regulates keratinocytes, which protect the skin against ultraviolet radiation via a dark pigment called melanin.

Neanderthal gene variants are also associated with a greater risk of sunburn in modern humans. Likewise, around 66% of Europeans carry a Neanderthal allele linked to a heightened risk of childhood sunburn and poor tanning ability.

Neanderthals had spent millennia at higher latitudes with less direct sun exposure, which is needed for vitamin D production. Therefore, changes to hair and skin biology may have allowed modern humans to quickly capitalize on lower levels of sunlight while still producing enough vitamin D to be healthy, John Capra, an evolutionary geneticist at Vanderbilt University, told Live Science.

"One of the cool things about interbreeding is that instead of waiting for new beneficial mutations to arise, which is a really slow process, you introduce a ton of genetic variation at once," essentially fast-tracking evolution, Huerta-Sanchez said.

Related: What's the difference between Neanderthals and Homo sapiens?

In addition, our ancestors had to adapt to colder Eurasian weather. To do so, they may have acquired Neanderthal genes that affected face shape. In a 2023 study, scientists discovered that modern humans inherited tall-nose genes from Neanderthals. A taller nose may have allowed more cold air to be heated to body temperature in the nose before reaching the lungs, suggested Kaustubh Adhikari, co-senior study author and a statistical geneticist at University College London.

The clock that makes our cells tick

Neanderthal DNA also may have helped H. sapiens adjust to the bigger differences in day and night length at northern latitudes.

Lingering Neanderthal genes affect our circadian clock, which regulates internal processes such as body temperature and metabolism. For instance, some early risers can thank Neanderthals for their circadian clock genes, Capra and colleagues found.

This may have helped our ancestors adapt to shorter winter days farther from the equator, Capra said.

"It seems like it's not that being a morning person is what matters," Capra said. "It's that that's a signal of how essentially flexible your clock is and how able it is to adapt to the variation in light-dark cycles with seasons," he said.

Our internal defenses

Many of the strongly retained Neanderthal genes are tied to immune function.

By the time H. sapiens arrived in Europe, Neanderthals had already spent hundreds of thousands of years fighting infections specific to Eurasia. By mating with Neanderthals, modern humans got an instant infusion of those infection-fighting genes.

"Those pieces of Neanderthal DNA, especially the immune ones, that were already adapted against pathogens that Neanderthals had been living with for a long time started to rise in frequency under natural selection in modern human populations," David Enard, an assistant professor of ecology and evolutionary biology at the University of Arizona, told Live Science.

While many of the ancestral pathogens that sickened ancient humans are lost to time, some of the Neanderthal genes that helped fight them off still work against modern pathogens. For example, a 2018 study by Enard and a colleague revealed that modern humans inherited Neanderthal DNA that helped them combat RNA viruses, a group that today includes the flu (influenza), HIV and hepatitis C.

Related: 10 unexpected ways Neanderthal DNA affects our health

The darker side of Neanderthal DNA

Some of the Neanderthal genes that once helped our ancestors may be harmful in the modern world.

For the most part, Neanderthal genes are not strongly expressed in the brain, which hints that they were strongly selected against during evolution. Neanderthal genes have been linked to mood disorders such as depression and to brain signaling pathways that make people more likely to become addicted to nicotine.

And even the immune boost from Neanderthals may have a downside. In 2016, scientists discovered that Neanderthal genes that prime the immune system to fight pathogens may also predispose people to allergic diseases. In addition, Neanderthal genes have been tied to a higher risk of developing autoimmune diseases, such as Graves' disease, caused by an overactive thyroid; and rheumatoid arthritis, which inflames the joints and even "Viking disease," in which one or more fingers become bent or frozen.

One Neanderthal gene variant may have made us more likely to have a severe case of COVID-19. That variant, found on chromosome 3, is found in half of South Asians and one-sixth of Europeans. But even there, the picture is complicated, as other Neanderthal genes, carried by up to half of people in Eurasia and the Americas, are associated with a reduced risk of severe COVID-19.

"Unfortunately, there are no diseases we can really say, or even traits in general, we can say, 'Oh, you can blame your Neanderthal DNA for that,'" Capra said.

That's especially true for some of the biggest health ailments, such as heart disease and cancer, where dozens or hundreds of genes, along with myriad environmental factors, affect your risk of disease.

What lies ahead

So how long will the traces of these long-lost humans linger in our genomes? Over hundreds of thousands of years, some of these Neanderthal fragments will gradually be eliminated from our genomes. Others will become firmly embedded, Akey said.

In the meantime, there's still much more to learn about how Neanderthals left their mark on us.

"Being able to leverage new genomic technology like CRISPR and gene editing is going to play an important role in understanding the actual underlying biology of how Neanderthal sequences contribute to human traits and diseases," Akey said.

Deciphering what these genes actually do could aid the development of treatments for certain conditions, he said.

And the gene flow wasn't one-way; scientists are also trying to determine how modern-human DNA may have influenced Neanderthals and are applying artificial intelligence (AI) methods to ancient genomes to create a more detailed picture of what our long-lost cousins were like.

Figuring out the role of Neanderthal DNA in our genomes does more than help us understand our health. These bits of DNA can provide clues as to what makes us unique, Sankararaman said.

"Neanderthal DNA entered our genomes at an important time in our history," Sankararaman said, when our ancestors were moving into new environments.

"By looking at the fate of these bits of DNA," he said, "we can hope to understand what were the functionally important regions in our genome over this period of time."

Editor's Note: This story was originally published in March 2024.

In science fiction films, nothing raises tension quite like the good guys' spaceship getting caught in an invisible tractor beam that allows the baddies to slowly reel them in. But what was once only a sci-fi staple could soon become a reality.

Scientists are developing a real-life tractor beam, dubbed an electrostatic tractor. This tractor beam wouldn't suck in helpless starship pilots, however. Instead, it would use electrostatic attraction to nudge hazardous space junk safely out of Earth orbit.

The stakes are high: With the commercial space industry booming, the number of satellites in Earth's orbit is forecast to rise sharply. This bonanza of new satellites will eventually wear out and turn the space around Earth into a giant junkyard of debris that could smash into working spacecraft, plummet to Earth, pollute our atmosphere with metals and obscure our view of the cosmos. And, if left unchecked, the growing space junk problem could hobble the booming space exploration industry, experts warn.

The electrostatic tractor beam could potentially alleviate that problem by safely moving dead satellites far out of Earth orbit, where they would drift harmlessly for eternity.

While the tractor beam wouldn't completely solve the space junk problem, the concept has several advantages over other proposed space debris removal methods, which could make it a valuable tool for tackling the issue, experts told Live Science.

Related: 11 sci-fi concepts that are possible (in theory)

A prototype could cost millions, and an operational, full-scale version even more. But if the financial hurdles can be overcome, the tractor beam could be operational within a decade, its builders say.

"The science is pretty much there, but the funding is not," project researcher Kaylee Champion, a doctoral student in the Department of Aerospace Engineering Sciences at the University of Colorado Boulder (CU Boulder), told Live Science.

Avoiding Disaster

The tractor beams depicted in "Star Wars" and "Star Trek" suck up spacecraft via artificial gravity or an ambiguous "energy field." Such technology is likely beyond anything humans will ever achieve. But the concept inspired Hanspeter Schaub, an aerospace engineering professor at CU Boulder, to conceptualize a more realistic version.

Schaub first got the idea after the first major satellite collision in 2009, when an active communications satellite, Iridium 33, smashed into a defunct Russian military spacecraft, Kosmos 2251, scattering more than 1,800 pieces of debris into Earth's orbit.

Related: How many satellites orbit Earth?

In the wake of this disaster, Schaub wanted to be able to prevent this from happening again. To do this, he realized you could pull spacecraft out of harm's way by using the attraction between positively and negatively charged objects to make them "stick" together.

Over the next decade, Schaub and colleagues refined the concept. Now, they hope it can someday be used to move dead satellites out of geostationary orbit (GEO) — an orbit around Earth's equator where an object's speed matches the planet's rotation, making it seem like the object is fixed in place above a certain point on Earth. This would then free up space for other objects in GEO, which is considered "prime real estate" for satellites, Schaub said.

How does it work?

The electrostatic tractor would use a servicer spacecraft equipped with an electron gun that would fire negatively charged electrons at a dead target satellite, Champion told Live Science. The electrons would give the target a negative charge while leaving the servicer with a positive charge. The electrostatic attraction between the two would keep them locked together despite being separated by 65 to 100 feet (20 to 30 meters) of empty space, she said.

Once the servicer and target are "stuck together," the servicer would be able to pull the target out of orbit without touching it. Ideally, the defunct satellite would be pulled into a "graveyard orbit" more distant from Earth, where it could safely drift forever, Champion said.

Related: 15 of the weirdest things we have launched into space

The electrostatic attraction between the two spacecraft would be extremely weak, due to limitations in electron gun technology and the distance by which the two would need to be separated to prevent collisions, project researcher Julian Hammerl, a doctoral student at CU Boulder, told Live Science. So the servicer would have to move very slowly, and it could take more than a month to fully move a single satellite out of GEO, he added.

That's a far cry from movie tractor beams, which are inescapable and rapidly reel in their prey. This is the "main difference between sci-fi and reality," Hammerl said.

Advantages and limitations

The electrostatic tractor would have one big advantage over other proposed space junk removal methods, such as harpoons, giant nets and physical docking systems: It would be completely touchless.

"You have these large, dead spacecraft about the size of a school bus rotating really fast," Hammerl said. "If you shoot a harpoon, use a big net or try to dock with them, then the physical contact can damage the spacecraft and then you are only making the [space junk] problem worse."

Scientists have proposed other touchless methods, such as using powerful magnets, but enormous magnets are both expensive to produce and would likely interfere with a servicer's controls, Champion said.

Related: How do tiny pieces of space junk cause incredible damage?

The main limitation of the electrostatic tractor is how slowly it would work. More than 550 satellites currently orbit Earth in GEO, but that number is expected to rise sharply in the coming decades.

If satellites were moved one at a time, then a single electrostatic tractor wouldn't keep pace with the number of satellites winking out of operation. Another limitation of the electrostatic tractor is that it would work too slowly to be practical for clearing smaller pieces of space junk, so it wouldn't be able to keep GEO completely free of debris.

Cost is the other big obstacle. The team has not yet done a full cost analysis for the electrostatic tractor, Schaub said, but it would likely cost tens of millions of dollars. However, once the servicer were in space, it would be relatively cost-effective to operate it, he added.

Next steps

The researchers are currently working on a series of experiments in their Electrostatic Charging Laboratory for Interactions between Plasma and Spacecraft (ECLIPS) machine at CU Boulder. The bathtub-sized, metallic vacuum chamber, which is equipped with an electron gun, allows the team to "do unique experiments that almost no one else can currently do" in order to simulate the effects of an electrostatic tractor on a smaller scale, Hammerl said.

Once the team is ready, the final and most challenging hurdle will be to secure funding for the first mission, which is a process they have not yet started.

Most of the mission cost would come from building and launching the servicer. However, the researchers would ideally like to launch two satellites for the first tests, a servicer and a target that they can maneuver, which would give them more control over their experiments but also double the cost.

Related: 10 stunning shots of Earth from space in 2022

If they can somehow wrangle that funding, a prototype tractor beam could be operational in around 10 years, the team previously estimated.

Is it viable?

While tractor beams may sound like a pipe dream, experts are optimistic about the technology.

"Their technology is still in the infancy stage," John Crassidis, an aerospace scientist at the University at Buffalo in New York, who is not involved in the research, told Live Science in an email. "But I am fairly confident it will work."

Removing space junk without touching it would also be much safer than any current alternative method, Crassidis added.

The electrostatic tractor "should be able to produce the forces necessary to move a defunct satellite" and "certainly has a high potential to work in practice," Carolin Frueh, an associate professor of aeronautics and astronautics at Purdue University in Indiana, told Live Science in an email. "But there are still several engineering challenges to be solved along the way to make it real-world-ready."

Scientists should continue to research other possible solutions, Crassidis said. Even if the CU Boulder team doesn't create a "final product" to remove nonfunctional satellites, their research will provide a stepping stone for other scientists, he added.

If they are successful, it wouldn't be the first time scientists turned fiction into fact.

"What is today's science fiction could be tomorrow's reality," Crassidis said.

In March 2019, researchers off the coast of southwestern Australia witnessed a gruesome scene: a dozen orcas ganging up on one of the biggest creatures on Earth to kill it. The orcas devoured huge chunks of flesh from the flanks of an adult blue whale, which died an hour later. This was the first-ever documented case of orca-on-blue-whale predation, but it wouldn't be the last.

In recent months, orcas (Orcinus orca) have also been spotted abducting baby pilot whales and tearing open sharks to feast on their livers. And off the coast of Spain and Portugal, a small population of orcas has begun ramming and sinking boats.

All of these incidents show just how clever these apex predators are.

"These are animals with an incredibly complex and highly evolved brain," Deborah Giles, an orca researcher at the University of Washington and the nonprofit Wild Orca, told Live Science. "They've got parts of their brain that are associated with memory and emotion that are significantly more developed than even in the human brain."

But the scale and novelty of recent attacks have raised a question: Are orcas getting smarter? And if so, what's driving this shift?

It's not likely that orcas' brains are changing on an anatomical level, said Josh McInnes, a marine ecologist who studies orcas at the University of British Columbia. "Behavioral change can influence anatomical change in an animal or a population" — but only over thousands of years of evolution, McInnes told Live Science.

Related: Scientists investigate mysterious case of orca that swallowed 7 sea otters whole

But orcas are fast learners, which means they can and do teach each other some terrifying tricks, and thus become "smarter" as a group. Still, some of these seemingly new tricks may in fact be age-old behaviors that humans are only documenting now. And just like in humans, some of these learned behaviors become trends, ebbing and flowing in social waves.

Frequent interactions with humans through boat traffic and fishing activities may also drive orcas to learn new behaviors. And the more their environment shifts, the faster orcas must respond and rely on social learning to persist.

Teaching hunting strategies 

There's no question that orcas learn from each other. Many of the skills these animals teach and share relate to their role as highly evolved apex predators.

Scientists described orcas killing and eating blue whales (Balaenoptera musculus) for the first time in a study published last year. In the months and years that followed the first attack in March 2019, orcas preyed on a blue whale calf and juvenile in two additional incidents, pushing the young blue whales below the surface to suffocate them.

This newly documented hunting behavior is an example of social learning, with strategies being shared and passed on from adult orcas to their young, Robert Pitman, a marine ecologist at Oregon State University's Marine Mammal Institute, told Live Science in an email. "Anything the adults learn will be passed along" from the dominant female in a pod to her offspring, he said.

Taking down a blue whale "requires cooperation and coordination," Pitman said. Orcas may have learned and refined the skills needed to tackle such enormous prey in response to the recovery of whale populations from whaling. This know-how was then passed on, until the orcas became highly skilled at hunting even the largest animal on Earth, Pitman said.

Old tricks, new observations 

Some of the gory behaviors researchers have observed recently may actually be long-standing habits.

For instance, during the blue whale attacks, observers noted that the orcas inserted their heads inside live whales' mouths to feed on their tongues. But this is probably not a new behavior — just a case of humans finally seeing it up close.

"Killer whales are like humans in that they have their 'preferred cuts of meat,'" Pitman said. "When preying on large whales, they almost always take the tongue first, and sometimes that is all they will feed on."

Tongue is not the only delicacy orcas seek out. Off the coast of South Africa, two males — nicknamed Port and Starboard — have, for several years, been killing sharks to extract their livers.

Although the behavior surprised researchers at first, it's unlikely that orcas picked up liver-eating recently due to social learning, Michael Weiss, a behavioral ecologist and research director at the Center for Whale Research in Washington state, told Live Science.

Related: Orcas attacked a great white shark to gorge on its liver in Australia, shredded carcass suggests

That's because, this year, scientists also captured footage of orcas slurping down the liver of a whale shark off the coast of Baja California, Mexico. The likelihood that Port and Starboard transferred their know-how across thousands of miles of ocean is vanishingly small, meaning liver-eating is probably a widespread and established behavior.

"Because there are more cameras and more boats, we're starting to see these behaviors that we hadn't seen before," Weiss said.

Sharing scavenging techniques 

Orcas master and share more than hunting secrets. Several populations worldwide have learned to poach fish caught for human consumption from the longlines used in commercial fisheries and have passed on this information.

In the southern Indian Ocean, around the Crozet Islands, two orca populations have increasingly scavenged off longlines since fishing in the region expanded in the 1990s. By 2018, the entire population of orcas in these waters had taught one another to feast on longline buffets, with whole groups that previously foraged on seals and penguins developing a taste for human-caught toothfish.

Sometimes, orcas' ability to quickly learn new behaviors can have fatal consequences. In Alaska, orcas recently started dining on groundfish caught by bottom trawlers, but many end up entangled and dead in fishing gear.

"This behavior may be being shared between individuals, and that's maybe why we're seeing an increase in some of these mortality events," McInnes said.

Playing macabre games 

Orcas' impressive cognitive abilities also extend to playtime.

Giles and her colleagues study an endangered population of salmon-eating orcas off the North Pacific coast. Called the Southern Resident population, these killer whales don't eat mammals. But over the past 60 years, they have developed a unique game in which they seek out young porpoises, with the umbilical cords sometimes still attached, and play with them to death.

Related: 'An enormous mass of flesh armed with teeth': How orcas gained their 'killer' reputation

There are 78 recorded incidents of these orcas tossing porpoises to one another like a ball but not a single documented case of them eating the small mammals, Giles said. "In some cases, you'll see teeth marks where the [killer] whale was clearly gently holding the animal, but the animal was trying to swim away, so it's scraping the skin."

The researchers think these games could be a lesson for young orcas on how to hunt salmon, which are roughly the same size as baby porpoises. "Sometimes they'll let the porpoise swim off, pause, and then go after it," Giles said.

Are humans driving orcas to become "smarter"? 

Humans may indirectly be driving orcas to become smarter, by changing ocean conditions, McInnes said. Orca raids on longline and trawl fisheries show, for example, that they innovate and learn new tricks in response to human presence in the sea.

Human-caused climate change may also force orcas to rely more heavily on one another for learning.

In Antarctica, for instance, a population of orcas typically preys on Weddell seals (Leptonychotes weddellii) by washing them off ice floes. But as the ice melts, they are adapting their hunting techniques to catch leopard seals (Hydrurga leptonyx) and crabeater seals (Lobodon carcinophaga) — two species that don't rely on ice floes as much and are "a little bit more feisty," requiring orcas to develop new skills, McInnes said.

While human behaviors can catalyze new learning in orcas, in some cases we have also damaged the bonds that underpin social learning. Overfishing of salmon off the coast of Washington, for example, has dissolved the social glue that keeps orca populations together.

"Their social bonds get weaker because you can't be in a big partying killer-whale group if you're all hungry and trying to search for food," Weiss said. As orca groups splinter and shrink, so does the chance to learn from one another and adapt to their rapidly changing ecosystem, Weiss said.

And while orcas probably don't know that humans are to blame for changes in their ocean habitat, they are "acutely aware that humans are there," McInnes said.

Luckily for us, he added, orcas don't seem interested in training their deadly skills on us.

From a distance, it might have looked like a small child was wending her way through the waving grass along a vast lake. But a closer look would have revealed a strange, in-between creature — a big-eyed imp with a small head and an apelike face who walked upright like a human.

She may have looked warily over her shoulder as she walked, on alert for saber-toothed cats or hyenas. She may have used her strong arms to climb the shrubby trees nearby, searching for fruit, eggs, or insects to eat. Or perhaps she simply rested on the shores of the croc-infested waters, gulping down water on a hot day.

She likely had no idea it was her last day on Earth.

Roughly 3.2 million years later, her skeleton was unearthed by paleoanthropologist Donald Johanson and his team on the International Afar Research Expedition.

The stunningly complete fossil was nicknamed "Lucy." And her remarkable species, Australopithecus afarensis, may have been our direct ancestor. Our discoveries about Lucy have transformed our understanding of humanity's tangled family tree.

Fifty years later, we know so much more about her species. In fact, anthropologists have learned so much about Lucy and her kind that we can now paint a picture of how she lived and died.

Her last day may have been filled with companionship, but it also entailed a relentless search for food. And it was likely dominated by the ever-present fear of predators.

"I suspect that the last day in her life was filled with danger," Johanson told Live Science.

Finding Lucy

The modern story of Lucy began on Nov. 24, 1974, in Hadar, Ethiopia. Johanson and then-graduate student Tom Gray stumbled upon a bone poking out of a gully. Following two weeks of careful excavation, their team recovered dozens of fossilized bones. Together, these bones made up 40% of the skeleton of a human ancestor, making it the most complete skeleton of an archaic human species that had ever been found.

Pamela Alderman, another member of the expedition, suggested the team nickname the skeleton Lucy, after the Beatles song "Lucy in the Sky with Diamonds."

"And it just became iconic," Johanson said, "a moniker that everybody knew."

Lucy’s discovery transformed the study of ancient human relatives.

"I was in high school when she was found," John Kappelman, a paleoanthropologist at the University of Texas at Austin, told Live Science. "It really did reset the way paleoanthropology worked."

Lucy's skeleton, along with subsequent discoveries of other fossils of her species, have given anthropologists a wealth of information about what is essentially the halfway point in human evolution. At 3.2 million years old, Lucy and her kind lived equidistant in time from our ape ancestors and contemporary humans.

"She's our touchstone," Jeremy DeSilva, a paleoanthropologist at Dartmouth College, told Live Science. "Everything sort of comes back to her as the reference point, and she deserves it."

"A lot like us"

One thing is fairly certain: Though there were some obvious differences, Lucy looked and acted a lot like us.

"If we saw her coming out of a grocery store today, we would recognize her as upright walking and some kind of human," Johanson said.

Although her strong arms and the shape of her finger bones suggest Lucy could climb trees, her pelvis and knees were clearly adapted to walking on two feet.

The size of Lucy's thigh bone also revealed that she was only about 42 inches (1.1 meters) tall and 60 to 65 pounds (27 to 30 kilograms) — about the size of a 6- or 7-year-old child today. And the eruption of her wisdom teeth showed that, although she was in her early teens when she died, she was a fully mature young adult.

"Australopithecus in general was maturing fast," DeSilva said, "and it makes sense if you're on a landscape full of predators." In species that are frequently prey, individuals that mature faster are more likely to pass on their genes. But australopithecines were unique—while their teeth and bodies matured quickly, their brains grew more slowly, telling us that they relied quite a bit on learning for survival, DeSilva said.

Her discovery also settled a debate that was raging in the early 1970s: Did our big brains evolve before we learned to walk upright? Lucy's head, which was not much bigger than a chimp's, showed the answer was no. Our ancestors became bipedal long before they evolved large brains.

Lucy's clan

Because her skeleton was found on its own, Lucy's "social life" is a little murkier than other parts of her daily life. But many researchers think she lived in a mixed-sex group of about 15 to 20 males and females, not unlike modern-day chimpanzees do.

And although there's no direct evidence, Lucy's skeletal maturity suggests she could have had a baby. Bringing that relatively large-headed newborn through her relatively narrow pelvis would have been challenging, which means she may have had the help of a primitive "midwife."

If Lucy had a baby, she also likely had a partner. Other A. afarensis fossils, such as those of Kadanuumuu, show male australopithecines were only slightly larger than females, which, in primates, usually corresponds to more monogamous pairings.

Lucy and her kind would have spent a significant amount of their time avoiding becoming another animal's lunch. "These small creatures would have been nice hors d'oeuvres for a sabertooth or a large cat or hyena," Johanson said.

Perhaps because of that omnipresent danger, the group likely relied on each other.

"I think they had each other's backs and helped each other out," DeSilva said, "especially when they were in dangerous situations."

A healed bone fracture seen in Kadanuumuu provides evidence that these primates cared for one another. Around 3.6 million years ago, this male australopithecine broke his lower leg. By the time he died, though, the break was fully healed.

"On that landscape with that many predators, no doctors, no hospitals, no casts, no crutches, how in the world do you survive if not for social assistance?" DeSilva said. "It's really strong evidence that they didn't leave each other for dead."

Lucy's last day

Lucy probably started her last day much like any other, waking up from the treetop nest made of branches and leaves where she slept, along with her group, before setting off to find food.

It's not clear whether she would have been alone or in a group when she left to forage; if she did have a baby, she may have carried it.

But there's no doubt that she would have spent a significant part of her day looking for food. She most likely ate a few staples, such as grasses, roots and insects, chemical elements in her tooth enamel showed. She may have happened upon the eggs of birds or turtles and promptly gobbled them up as tasty, protein-rich treats. And if she was lucky enough to come across a carcass of a large mammal, such as an antelope, that hadn't been picked clean, she and her troop mates may have pulled the flesh from the bone, using large rocks.

"They can't afford to be picky eaters as these slow bipeds in a dangerous environment," DeSilva said. "They're eating everything they can get their hands on."

However, there's no evidence that Lucy’s species used fire to cook any of their food.

Death at the water's edge

In the past 50 years, we've created a picture of Lucy's last moments. It's not clear exactly why she was by the lake; maybe she was thirsty, or perhaps it was a great spot to look for food.

But there are two main theories for how she died.

"Perhaps she was down there at the water and — bam! — a crocodile comes out," Johanson said. "Crocodiles are incredibly fast, and it's a dangerous place if you're a little creature" like Lucy.

Johanson found one carnivore tooth mark on Lucy's pelvis, and it had not healed, meaning it occurred around the time of her death. Although the animal that made the mark has not been conclusively identified, "we know that australopithecines were preyed upon because there are a number of examples," Johanson said.

In 2016, Kappelman and his colleagues put forward an alternate ending for Lucy: a catastrophic fall from a tree.

Based on high-resolution CT scans and 3D reconstructions of Lucy's skeleton, Kappelman identified fractures in her right shoulder, ribs and knees that were unlike the typical fracturing that occurs in fossils crushed under the weight of dirt and rocks for millions of years.

"Something traumatic happened here during life," Kappelman said.

The kinds of fractures Lucy suffered are consistent with a fall from a considerable height, perhaps from a tall tree in which she was foraging for food.

"She hit on her feet and then her hands, which meant she was conscious when she hit the ground," Kappelman said. "I don't think she survived very long."

It's not clear whether she was alone when she died. But even if she was with others of her kind, they likely wouldn't have done much with her body.

There's no evidence that A. afarensis "bodies were treated any differently than any other animal," DeSilva said. "Maybe there was some curiosity around it, and then they carried on."

Primate researchers have documented other species' curiosity about inanimate bodies. For example, chimpanzees often care for the body for a few hours or days after death, sometimes guarding the body.

Lucy's group may have done the same for her until her body was naturally buried, which would have happened quite rapidly, perhaps by a flood or mudslide.

In the end, though, "we know very little about how any of these creatures died," Johanson said.

Lucy lives on

Thanks to Johanson's 1974 discovery of Lucy — as well as other important findings, like the "First Family" and the footprints at Laetoli in Tanzania — we now know quite a lot about A. afarensis.

"It was a highly successful species that was comfortable in lots of different habitats," Johanson said; A. afarensis fossils have been found in Kenya in addition to Ethiopia and Tanzania. "From an evolutionary perspective, her species was highly adaptable," he said.

Lucy has had a broad impact on the field of anthropology.

"The discovery of Lucy really hit the start button for looking in older and older sediments in Africa," Kappelman said. As a result, we have found numerous ancient hominin species and now have 50 years' worth of fossil evidence that human evolution was messy and complex.

Lucy was the only human ancestor discovered at Hadar. But a couple dozen miles away at Woranso-Mille, a paleontological site in Ethiopia, Yohannes Haile-Selassie, director of the Institute of Human Origins at Arizona State University, and his colleagues have found evidence of a strange land inhabited by multiple humanlike species between 3.8 million and 3.3 million years ago. For instance, Lucy's kind coexisted alongside another ancient relative, A. anamensis.

Would they have been friends, enemies, competitors or something in between? Right now, anthropologists still have little idea what this landscape teeming with ancient hominins would have looked like.

But perhaps 50 years from now, we'll have a better picture of how Lucy's kind interacted with these other ancient hominins. Even then, Lucy will likely remain one of the most famous fossils of all time.

"I like to think all fossils are pretty special," DeSilva said, "but there's nothing like Lucy."

Editor's note: This article was originally published in November, 2024 as part of a special package written for the 50th anniversary of the discovery of a 3.2 million-year-old A. afarensis fossil (AL 288-1), nicknamed "Lucy."

We've watched it speed through the solar system using the most powerful telescopes in human history. We've studied its light with probes whipping around the sun and robots marooned on Mars. Countless eyes watched it make its closest approach to Earth on Dec. 19 — and yet, for all of this, the interstellar comet 3I/ATLAS remains little more than a blur of gas, shrouded in mystery.

Since its discovery in early July, 3I/ATLAS has been studied more enthusiastically than practically any other celestial object in recent memory. Still, for all its fame, much remains unknown about it. The comet’s origins, from somewhere far across our galaxy, may never be known. Its true age, size, composition, and shape are also poorly constrained.

But how can we learn more about this alien interloper — or indeed, the next one — when we’re already studying it with everything we’ve got?

Some scientists are proposing a bold solution: We have to "intercept" it with a spacecraft.

Doing so would not only help us to better understand its key characteristics but also photograph its surface and potentially collect our first-ever interstellar samples, which could help reveal how alien exoplanets form, how common our type of solar system is and maybe even help answer the question of whether or not we are alone in the universe.

"We only have one shot at this object and then it's gone forever," Darryl Seligman, an astronomer at Michigan State University and the lead author of the first paper published about 3I/ATLAS, previously told Live Science. "So we want as much information from all of our observatories as we can possibly get."

Alien interlopers

On July 1, astronomers at the Asteroid Terrestrial-impact Last Alert System (ATLAS) revealed they had spotted a mysterious object traveling toward us from beyond Jupiter, at more than 130,000 mph (210,000 km/h). ATLAS, which automatically scans the skies using telescopes in Hawaii, Chile and South Africa, was hunting for potential threats to Earth. It found something else entirely.

Less than 24 hours later, NASA confirmed that the speeding blur of light was an interstellar object — an alien asteroid or comet that originated outside the solar system — and named it 3I/ATLAS. It was only the third-ever detection of an interstellar object in our solar system, after the anomalous space rock 'Oumuamua in 2017 and Comet 2I/Borisov in 2019.

Despite the rapid spread of unfounded theories that the object could be an alien probe, early observations confirmed that 3I/ATLAS is a comet — potentially the oldest of its kind ever seen — that likely originated from the Milky Way's "frontier" region.

Interstellar visitors like this are exciting to astronomers because they are one of the few opportunities we have to explore neighboring star systems, which would take generations and the invention of sci-fi technology to reach aboard a spacecraft.

"ISOs are relics from planetary formation, so studying these objects and comparing them to what we have closer to us [could] lead to an interesting view of how other planetary systems in the galaxy formed," Pedro Bernardinelli, a planetary scientist at the University of Washington's DiRAC Institute, told Live Science in an email.

But our Earth-based observatories, and even orbiting spacecraft such as the James Webb Space Telescope (JWST), can only tell us very rough information like general size, shape and composition. To really reveal ISO secrets, we will need to get much, much closer — possibly even close enough to grab a fragment.

Doing so won't be easy, but given the valuable insights it could reveal about the star systems beyond our own, it would be well worth the effort, experts say.

"Each one of these ISOs is a little piece of low-hanging fruit from a tree that can tell us a great deal about the trees growing in some other neighborhood," Wesley Fraser, an astronomer with the National Research Council Canada, previously told Live Science.

Giving chase

But the time to catch this speeding comet is fast approaching. 3I/ATLAS is now reaching its closest point to Earth, around 168 million miles (270 million km) miles away. From there it will move quickly away from us and will likely be beyond Neptune within another year.

Because it is now too late to intercept 3I/ATLAS within the inner solar system, most researchers agree that there is now only one viable option to study this object: to chase it down as it leaves the solar system.

This would require the spacecraft to carry out what researchers call "Oberth maneuvers," where a probe is gravitationally slingshotted around massive objects, such as the sun, to pick up enough speed to allow it to catch up to and intercept an ISO at a specific point along its predicted trajectory.

This idea was first proposed in 2022 to catch up with the first known interstellar object, 'Oumuamua. The plan, dubbed Project Lyra, was to launch a probe in 2028 that would intercept and investigate that object, after completing an Oberth maneuver around Jupiter.

But this chaser method has a huge limitation: Scientists would need to wait decades for data to come back. For example, if Project Lyra launched a spacecraft in 2030, it would not intercept 'Oumuamua until 2052 at the earliest, Adam Hibberd, a researcher with the U.K.-based nonprofit Initiative for Interstellar Studies (I4IS) who worked on Project Lyra, told Live Science.

So far, Project Lyra has not moved past the planning stage — making a 2028 launch highly unlikely — but the project could still reach 'Oumuamua if launched in 2030 or 2033, Hibberd said. This means we would likely still have plenty of time to chase down 3I/ATLAS, if we want to.

Future propulsion methods, such as a solar sail, could drastically cut the travel time of missions like this from decades down to just a few years, he added. But these technologies are decades away from becoming a reality themselves.

Playing "hide-and-seek"

But given that 3I/ATLAS will be very hard to chase down, some astronomers argue that we shouldn't bother hunting it. Rather we should prepare to intercept the next interesting ISO.

By launching an interceptor spacecraft and parking it in a gravitationally stable position around Earth, known as a Lagrange point, we could, in theory, be ready to quickly intercept a passing object, they argue.

This idea, also first proposed in 2022, has been dubbed the "hide-and-seek" approach. However, unlike Project Lyra, it is much closer to becoming a reality.

The European Space Agency (ESA) is preparing the Comet Interceptor mission, which is currently scheduled to launch in 2029, on board the same rocket as ESA's Ariel space telescope, said Colin Snodgrass, an astronomer at the University of Edinburgh in Scotland who specializes in comets and was the deputy project investigator on the proposal for this mission.

The Comet Interceptor probe isn't specifically aimed at interstellar visitors. Instead, it's designed to hunt nonperiodic comets like Comet Lemmon, which has been visible in the night sky, alongside 3I/ATLAS, in recent months. These comets drift toward the sun every few hundred or thousand years and have poorly defined orbital pathways around the sun.

When ESA researchers spot a comet they can reach, they will "fire the rockets, get to the right place in space to cross the path of the comet and have this fast flyby encounter, where we go shooting past the comet, getting as much data as we can," Snodgrass told Live Science.

And while the mission is not designed to study interstellar objects, the project will be perfectly placed to intercept them.

"The whole science team is very much in agreement that if an interstellar object was to pop up, we wouldn't let that opportunity go by," Snodgrass said.

The main advantage of the hide-and-seek approach is that we wouldn't have to wait decades for a probe to catch up to its target. Additionally, we'd be reaching it at the best time to study it. That's because interstellar comets, like 3I/ATLAS, soak up more solar radiation when in the inner solar system — which, in turn, means they give off more light, gas and dust, giving us a better chance to learn about their composition.

However, a hide-and-seek mission might not be able to catch all the objects we care about. For example, ESA's Comet Interceptor probe would have been unlikely to reach 3I/ATLAS, had it been in orbit when the ISO was first discovered, because the comet was too far away from us, a recent study from Snodgrass and others found.

Collision course

A major limitation of both the chaser and hide-and-seek missions is that ISOs travel too fast for their respective spacecraft to travel alongside, or rendezvous with, these objects.

This makes it "almost impossible" for the probes to directly obtain samples from the objects' surfaces as NASA did during its OSIRIS-REx mission, which successfully landed a probe on the asteroid Bennu in 2020 and collected samples that were later returned to Earth, Hibberd said. Due to fuel limitations, it is also unlikely that these samples could be easily returned to Earth, especially during a chaser mission, he added.

However, there is a third option that could yield valuable interstellar samples: the "impactor" method.

Similar to NASA's Double Asteroid Redirection Test (DART) mission, which successfully deflected the asteroid Dimorphos after smashing into the space rock in 2022, an interceptor probe could also be sent to crash into an ISO, Hibberd suggested. While this probe would be destroyed, a second spacecraft could be deployed to analyze the debris field and potentially even collect leftover fragments of the alien object, he added.

But an impactor mission would need to overcome serious technical challenges. First, ISOs travel much faster than solar system objects, like Dimorphos, meaning it's more difficult to smash them apart. Second, this method would likely work only on an asteroid, not on comets, which have hard, icy shells. And third, a collision could accidentally send chunks of debris on a collision course with Earth, like DART did. As a result, most of the experts who talked to Live Science, including Hibberd, agreed that it is probably too risky to attempt an impactor mission until more research has been done on the subject.

The perfect mission

If money were no object, we could pursue all of these options. But if an agency like NASA has the budget for only one such mission, which one should be selected?

A chaser mission would allow astronomers to target a specific object they know they want to study, while a hide-and-seek mission would be limited to objects that happened to pass nearby. On the other hand, the hide-and-seek mission could reliably predict objects' locations in the inner solar system, whereas the chaser method would target objects in the dark, more chaotic outer solar system, where it would be harder to find and photograph them, Snodgrass said.

Another issue is that signals from a more distant chaser mission would take longer to send and receive, so mission operators would be unable to monitor and adjust an ISO flyby in real time or fix technical difficulties easily — a difficulty NASA faces with its distant Voyager probes, Snodgrass said.

There is also the matter of money. Project Lyra would likely cost the same as NASA's New Horizons mission, which flew by Pluto in 2015 and cost at least $700 million, Hibberd said. Meanwhile, ESA's Comet Interceptor mission has a budget of around $150 million, Snodgrass said.

As a result, most researchers who spoke to Live Science agreed that a hide-and-seek interceptor would likely be the best way of studying an ISO up close.

But if this is the method we end up using, how should we design the resulting spacecraft to maximize its chances of collecting useful data?

While ESA's Comet Interceptor is relatively inexpensive, a dedicated ISO interceptor mission — with a bigger budget — would allow us to launch a faster probe that could carry more fuel and thus travel farther. However, the craft doesn't need to be fancy.

A "fairly stripped-back" probe with a decent camera and a few spectrographs, capable of analyzing the light given off by the different gases, would be more than enough to collect sufficient data from any flyby, Snodgrass said.

If the probe were intercepting a comet, and not an asteroid, it could also be fitted with a device to catch specks of dust from the comet's coma or tail during a superclose approach, just as NASA's Stardust probe did with "Comet Wild 2" in 2004.

Assuming that the interceptor hasn't depleted its fuel reserves and can be returned to Earth, this may be the only reliable way of actually getting our hands on interstellar samples, Snodgrass said.

To intercept or not to intercept

Once the "perfect" interceptor is in position around Earth, researchers will have to choose which ISO to go after. And because any spacecraft is unlikely to be reusable, it may get only one shot at picking the right target.

We may soon be spoiled for choice. ISOs may be far more common than we realize. "There are likely thousands of other ISOs in the solar system right now," Fraser said. "We just can't see them because they are too faint, too far and too fast."

But thanks to the newly operational Vera C. Rubin Observatory in Chile, which is designed to spot more small and dim objects in the outer solar system, we are likely to find many more ISOs in the coming decades and, more importantly, spot them much earlier on their journey toward us, which would give us a better chance of studying them.

The first thing to consider is whether to go after an asteroid or a comet. Because comets become more active near the sun and present the most likely route for collecting interstellar samples, they would likely take priority, Snodgrass said.

The next consideration would be the target's distance from Earth. As we have already seen, ESA's Comet Interceptor may have struggled to reach 3I/ATLAS on its journey through the inner solar system. Therefore, it might pay to wait for an ISO that is on a favorable trajectory relative to Earth.

In October 2024, news broke that Facebook parent company Meta had cracked an "impossible" problem that had stymied mathematicians for a century.

In this case, the solvers weren't human.

An artificial intelligence (AI) model developed by Meta determined whether solutions of the equations governing certain dynamically changing systems — like the swing of a pendulum or the oscillation of a spring — would remain stable, and thus predictable forever.

The key to the problem was finding Lyapunov functions, which determine the long-term stability of these systems.

Meta's work made headlines and raised a possibility once considered pure fantasy: that AI could soon outperform the world's best mathematicians by cracking math's marquee "unsolvable" problems en masse.

After looking under the hood, however, mathematicians were less impressed. The AI found Lyapunov functions for 10.1% of randomly generated problems posed to it. This was a substantial improvement over the 2.1% solved by previous algorithms, but it was by no means a quantum leap forward. And the model needed lots of hand-holding by humans to come up with the right solutions.

A similar scenario played out earlier this year, when Google announced its AI research lab DeepMind had discovered new solutions to the Navier-Stokes equations of fluid dynamics. The solutions were impressive, but AI was still some distance from solving the more general problem associated with the equations, which would garner its solvers the $1 million Millennium Prize.

Beyond the hype, just how close is AI to replacing the world's best mathematicians? To find out Live Science asked some of the world's best mathematicians.

While some experts were dubious about AI’s problem solving abilities in the short term, most noted that the technology is developing frighteningly fast. And some speculated that not so far into the future, AI may be able to solve hard conjectures — unproven mathematical hypotheses — at a massive scale, invent new fields of study, and tackle problems we never even considered.

"I think what's going to happen very soon — actually, in the next few years — is that AIs become capable enough that they can sweep through the literature at the scale of thousands — well, maybe hundreds, tens of thousands of conjectures," UCLA mathematician Terence Tao, who won the Fields Medal (one of mathematics' most prestigious medals) for his deep contributions to an extraordinary range of different mathematical problems, told Live Science. "And so we will see what will initially seem quite impressive, with thousands of conjectures suddenly being solved. And a few of them may actually be quite high-profile ones."

From games to abstract reasoning

To understand where we are in the field of AI-driven mathematics, it helps to look at how AI progressed in related fields. Math requires abstract thinking and complex multistep reasoning. Tech companies made early inroads into such thinking by looking at complex, multistep logical games.

In the 1980s, IBM algorithms began making progress in games like chess. It's been decades since IBM's Deep Blue beat what was then the world's best chess player, Garry Kasparov, and about a decade since Alphabet's DeepMind defeated the period's best Go player, Lee Sedol. Now AI systems are so good at such mathematical games that there's no point to these competitions because AI can beat us every time.

But pure math is different from chess and Go in a fundamental way: Whereas the two board games are very large but ultimately constrained (or, as mathematicians would say, "finite") problems, there are no limits to the range, depth and variety of problems mathematics can reveal.

In many ways, AI math-solving models are where chess-playing algorithms were a few decades ago. "They're doing things that humans know how to do already," said Kevin Buzzard, a mathematician at Imperial College London.

"The chess computers got good, and then they got better and then they got better," Buzzard told Live Science. "But then, at some point, they beat the best human. Deep Blue beat Garry Kasparov. And at that moment, you can kind of say, 'OK, now something interesting has happened.'"

That breakthrough hasn't happened yet for math, Buzzard argued.

"In mathematics we still haven't had that moment when the computer says, 'Oh, here's a proof of a theorem that no human can prove,'" Buzzard said.

Mathematical genius?

Yet many mathematicians are excited and impressed by AI's mathematical prowess. Ken Ono, a mathematician at the University of Virginia, attended this year's "FrontierMath' meeting organized by OpenAI. Ono and around 30 of the world's other leading mathematicians were charged with developing problems for o4-mini — a reasoning large language model from OpenAI — and evaluating its solutions.

After witnessing the heavily human-trained chatbot in action, Ono said, "I've never seen that kind of reasoning before in models. That's what a scientist does. That's frightening." He argued that he wasn't alone in his high praise of the AI, adding that he has "colleagues who literally said these models are approaching mathematical genius."

To Buzzard, these claims seem far-fetched. "The bottom line is, have any of these systems ever told us something interesting that we didn't know already?" Buzzard asked. "And the answer is no."

Rather, Buzzard argues, AI's math ability seems solidly in the realm of the ordinary, if mathematically talented, human. This summer and last, several tech companies' specially trained AI models attempted to answer the questions from the International Mathematical Olympiad (IMO), the most prestigious tournament for high school "mathletes" around the world. In 2024, Deepmind's AlphaProof and AlphaGeometry 2 systems combined to solve four of the six problems, scoring a total of 28 points — the equivalent of an IMO silver medal. But the AI first required humans to translate the problems into a special computer language before it could begin work. It then took several days of computing time to solve the problems — well outside the 4.5-hour time limit imposed on human participants.

This year's tournament witnessed a significant leap forward. Google's Gemini Deep Think solved five of the six problems well within the time limit, scoring a total of 35 points. This is the sort of performance that, in a human, would have been worthy of a gold medal — a feat achieved by less than 10% of the world's best math students.

Research-level problems

Although the most recent IMO results are impressive, it's debatable whether matching the performance of the top high school math students qualifies as "genius-level."

Another challenge in determining AI's mathematical prowess is that many of the companies developing these algorithms don't always show their work.

"AI companies are sort of shut. When it comes to results, they tend to write the blog post, try and go viral and they never write the paper anymore," Buzzard, whose own research lies at the interface of math and AI, told Live Science.

However, there's no doubt that AI can be useful in research-level mathematics.

In December 2021, University of Oxford mathematician Marc Lackenby's research with DeepMind was on the cover of the journal Nature.

Lackenby's research is in the area of topology which is sometimes referred to as geometry (the maths of shapes) with play dough. Topology asks which objects (like knots, linked rings, pretzels or doughnuts) keep the same properties when twisted, stretched or bent. (The classic math joke is that topologists consider a doughnut and a coffee cup to be the same because both have one hole.)

Lackenby and his colleagues used AI to generate conjectures connecting two different areas of topology, which he and his colleagues then went on to try to prove. The experience was enlightening.

It turned out that the conjecture was wrong and that an extra quantity was needed in the conjecture to make it right, Lackenby told Live Science.

Yet the AI had already seen that, and the team "had just ignored it as a bit of noise," Lackenby said.

Can we trust AI at the frontier of math?

Lackenby's mistake had been not to trust the AI enough. But his experience speaks to one of the current limitations of AI in the realm of research mathematics: that its outputs still need human interpretation and can't always be trusted.

"One of the problems with AI is that it doesn't tell you what that connection is," Lackenby said. "So we have to spend quite a long time and use various methods to get a little bit under the hood."

Ultimately, AI isn't designed to get the "right" answer; it's trained to find the most probable one, said Neil Saunders, a mathematician who studies geometric representation theory at City St George's, University of London and the author of the forthcoming book "AI (r)Evolution" (Chapman and Hall, 2026), told Live Science.

"That most probable answer doesn't necessarily mean it's the right answer," Saunders said.

AI's unreliability means it wouldn't be wise to rely on it to prove theorems in which every step of the proof must be correct, rather than just reasonable.

"You wouldn't want to use it in writing a proof, for the same reason you wouldn't want ChatGPT writing your life insurance contract," Saunders said.

Despite these potential limitations, Lackenby sees AI's promise in mathematical hypothesis generation. "So many different areas of mathematics are connected to each other, but spotting new connections is really of interest and this process is a good way of seeing new connections that you couldn't see before," he said.

The future of mathematics?

Lackenby's work demonstrates that AI can be helpful in suggesting conjectures that mathematicians can then go on to prove. And despite Saunders' reservations, Tao thinks AI could be useful in proving existing conjectures.

The most immediate payoff might not be in tackling the hardest problems but in picking off the lowest-hanging fruit, Tao said.

The highest-profile math problems, which "dozens of mathematicians have already spent a long time working on — they're probably not amenable to any of the standard counterexamples or proof techniques," Tao said. "But there will be a lot that are."

Tao believes AI might transform the nature of what it means to be a mathematician.

"In 20 or 30 years, a typical paper that you would see today might indeed be something that you could automatically do by sending it to an AI," he said. "Instead of studying one problem at a time for months, which is the norm, we're going to be studying 10,000 problems a year … and do things that you just can't dream of doing today."

Rather than AI posing an existential threat to mathematicians, however, he thinks mathematicians will evolve to work with AI.

"We've had situations in the past where entire fields of mathematics became basically solvable by computer," Tao said. At one point, we even had a human profession called a "computer," he added. That job has disappeared, but humans just moved on to harder problems. "It didn't mean mathematics died," Tao said.

Andrew Granville, a professor of number theory at the University of Montreal, is more circumspect about the future of the field. "My feeling is that it's very unclear where we're going," Granville told Live Science. "What is clear is that things are not going to be the same. What that means in the long term for us depends on our adaptability to new circumstances."

Lackenby similarly doesn't think human mathematicians are headed for extinction.

While the precise degree to which AI will infiltrate the subject remains uncertain, he's convinced that the future of mathematics is intertwined with the rise of AI.

"I think we live in interesting times," Lackenby said. "I think it's clear that AI will have an increasing role in mathematics."

In 1987, a worker lit a cigarette by a new water well near the village of Bourakebougou in Mali.

But as he did, an explosion reverberated inside the well. We now know this was due to previously undetected clouds of flammable hydrogen wafting from a gas reservoir beneath the hole.

The well was plugged and temporarily abandoned. Then, in 2011, the oil and gas company Petroma (which later became Hydroma) uncemented it to examine whether they could extract hydrogen for profit. By 2012, the company had developed the well to make electricity for Bourakebougou, and the village still relies on this hydrogen for power today.

Bourakebougou's well is the world's first and only productive hydrogen well. Mixed with oxygen in fuel cells, hydrogen — the smallest and simplest molecule in existence — can generate electricity without greenhouse gas emissions and with only heat and water as byproducts. This makes hydrogen a clean source of energy, and demand for it is expected to rise fivefold by 2050 to produce microelectronics, supply industry, and power vehicles and buildings.

Hydrogen is lighter than air and very reactive, so scientists long thought it didn't accumulate inside Earth's crust in the same way fossil fuels do. But the discovery in Bourakebougou, along with more recent finds, has completely shifted this paradigm.

Resource exploration companies are now rushing to find reservoirs of natural hydrogen, also known as "gold" hydrogen. To help them, scientists have identified the key "ingredients" needed to form such accumulations. And thanks to this knowledge, techniques to boost or mimic natural hydrogen generation that were once considered impracticable are gaining traction, experts told Live Science.

"We just keep finding more and more the more we start looking for it," Geoffrey Ellis, a petroleum geochemist with the U.S. Geological Survey, told Live Science.

Paradigm shift

Hydrogen is a source of energy, but it is also a critical component of fertilizer, refined oil and rocket fuel. Industry produces almost all of its hydrogen by heating natural gas with steam to form a mixture of hydrogen and carbon monoxide from which hydrogen can be extracted.

This method makes "gray" hydrogen, and it pumps about 1 billion tons (920 million metric tons) of carbon dioxide into the atmosphere every year — equivalent to 2.4% of global annual emissions. In theory, renewable energies can replace natural gas to generate "green" hydrogen, while "blue" hydrogen is made from fossil fuels but with carbon capture, meaning carbon doesn't enter the atmosphere. But these collectively make up a tiny fraction of hydrogen production worldwide.

"Hydrogen is a clean source of energy, but how you get your hydrogen is critical," Chris Ballentine, a professor of geochemistry at the University of Oxford, told Live Science.

However, a new source of hydrogen could slash the industry's carbon footprint, as it turns out that huge quantities of hydrogen can accumulate belowground. Scientists have long known that rocks in Earth's crust produce hydrogen, but experts previously concluded that the gas couldn't collect in reservoirs because only tiny concentrations of it were being found in oil and gas wells.

The discovery in Mali toppled that theory. Researchers realized that the places where companies drill for oil and gas are not the best places to find hydrogen.

Massive reservoirs, waiting to be found

The Mali discovery has kicked off a worldwide hunt for hydrogen reservoirs. But before geologists initiate costly exploration projects, they need a sense of just how much hydrogen might be lurking underground.

New estimates suggest it's a staggering amount. Earth's continental crust has produced enough hydrogen over the past 1 billion years to meet society's current energy needs for 170,000 years, a recent review by Ballentine and his colleagues found. Though much of this hydrogen has escaped into the atmosphere, the figure is "a starting point for realizing that the hydrogen generation in the crust is significant," Ballentine said.

Other estimates double the figure in the Ballentine paper. Ophiolites are chunks of oceanic crust that have been thrust onto the continental crust, and some estimates suggest these ocean-crust remnants may produce as much hydrogen as the continental crust does, Ballentine said.

But how much of this hydrogen is left in Earth's crust? In 2024, Ellis and his colleagues calculated that the planet holds 6.2 trillion tons (5.6 trillion metric tons) of hydrogen, or about 26 times the amount of oil known to be left in the ground. Where these hydrogen stocks are located is largely unknown. Most are likely too deep or too far offshore to be accessed, and some reservoirs might be too small to be worth extracting — but the researchers emphasized that just 2% of the total hydrogen could supplant our current fossil fuels for 200 years.

"The potential that's down there is quite, quite large," Ellis said. What's more, natural hydrogen, unlike the type made via industrial processes, comes with built-in storage because it sits in Earth's crust. It also has a much smaller carbon footprint than manufactured hydrogen, with emissions coming only from extraction, Ellis said.

The ingredients

In January 2025, Ellis and his colleagues published a map showing where hydrogen reservoirs might exist in the lower 48 U.S. states. The researchers used gravity and magnetic signal data to estimate the composition of rocks throughout Earth's crust and determine where hydrogen may have migrated underground.

"This was the first time that anyone had attempted to do this type of mapping exercise," Ellis said.

The researchers estimated the likelihood of productive hydrogen reservoirs, known as prospectivity, based on six geological requirements that make and trap hydrogen in Earth's crust. On the map, prospectivity ranges from 0 to 1, with 0 meaning there is likely no hydrogen and 1 indicating hydrogen is very likely present.

To form a hydrogen reservoir, the first and second requirements are that a region must have abundant groundwater and hydrogen-producing rocks. The water requirement limits hydrogen production to the top 10 miles (16 kilometers) of the crust, Oliver Warr, an assistant professor of geochemistry at the University of Ottawa, told Live Science.

The best hydrogen-producing rocks are iron-rich rocks, which generate hydrogen through "hydration reactions," where water reacts with the rocks. Other good sources of hydrogen are uranium- and thorium-rich rocks, which produce alpha particles as the radioactive elements decay. These alpha particles can then split water into oxygen and hydrogen — a process known as radiolysis, Warr said.

Iron-rich rocks include basalt and gabbro. Earth's mantle, the layer beneath the crust, heats groundwater, producing steam that reacts with iron and generates hydrogen. Uranium- and thorium-rich rocks include granites, and these can trigger the radiolysis of water.

The third requirement is that the source rocks be very, very hot — between 480 and 570 degrees Fahrenheit (250 to 300 degrees Celsius), which guarantees rapid rates of reaction, Ellis said.

Fourth, the region must have reservoir rocks that can hold the hydrogen after it is produced and migrates through the crust. Reservoir rocks are typically porous sandstones, but other types of rock can also work if they are highly fragmented, Ellis said.

The fifth criterion to form a hydrogen reservoir is an impermeable "seal" to trap the gas inside the reservoir. "A thing like a shale, or maybe a salt, would be really ideal to be sitting on top of that porous rock," Ellis said. Crucially, the seal must exist when the hydrogen is produced, or else the gas escapes into the atmosphere, he said.

The sixth and final condition is that there must be minimal microbial activity where hydrogen is generated and accumulates, because microbes consume hydrogen, Warr said.

These six conditions, or ingredients, occur across all continents, Ballentine said. Currently, hydrogen companies are drilling exploratory wells mostly on the Midcontinent Rift — where North America started, but ultimately failed, to split apart 1 billion years ago — which is abundant in iron-rich rocks.

Looking ahead

Researchers are also investigating hydrogen deposits in Oman, where there are ophiolites. University of Colorado geologists are running a pilot project in the country to test the feasibility of "stimulated hydrogen" production, Ellis said.

Stimulated hydrogen production takes inspiration from what scientists have learned about the geology that makes and accumulates hydrogen. It involves injecting water into Earth's crust to kick-start either hydration reactions or radiolysis.

One year ago, people in the hydrogen industry were skeptical that stimulated hydrogen production would ever materialize, Ellis said. But now, "I've seen a big shift," he said.

If we can find natural hydrogen and extract it, the gas could reduce emissions across a wide range of sectors. For example, abundant hydrogen is found in mines, because this is where humans drill deepest into the crust, so the gas could power mining operations, Warr said.

Natural hydrogen could also slash emissions from industries such as fertilizer manufacturing. "If we can replace hydrogen generated from hydrocarbons with clean hydrogen, then we can very rapidly make a massive difference," Ballentine said.

Natural hydrogen won't solve the climate crisis, but it can mitigate some of the risks. "It needs to be one of many strategies," Warr said. "We just need to understand the true potential and how it can best be capitalized on."

Some of the key considerations for companies are whether the benefits of developing natural hydrogen reservoirs when we find them would justify the cost of building production plants on-site, or shipping the gas to the industries that need it.

"If you're remote and you find a really big gas field, it still may not be worthwhile producing it, because the costs of getting hydrogen to market are too great," Ballentine said. "There's a trade-off."

But overall, experts are optimistic. "There have been, I think, over a dozen wells that have been drilled now in the U.S.," Ellis said. "They've found a lot of hydrogen."

In 1903, residents of the small town of Dexter, Kansas, gathered to celebrate a newly drilled natural gas well.

Crowds flocked to the well for the lighting of the escaping gas, which officials had said would produce "a great pillar of flame." But when they rolled a burning bale of hay onto the well, nothing happened.

An analysis in 1905 revealed that most of the gas was nonflammable nitrogen, just 15% was methane — and a little less than 2% was a colorless, odorless, elusive element that scientists had discovered only a few decades earlier. This marked the first discovery of helium in a natural gas field.

Helium is the second-most-abundant element in the universe after hydrogen, yet it is scarce on Earth. Helium does not react with other elements, and it mostly accumulates in Earth's crust as radioactive uranium and thorium decay, which takes billions of years. Because it takes much longer to replenish helium reserves than it does to deplete them, helium is a nonrenewable resource.

The discovery in Dexter paved the way for the helium industry, which mushroomed during World War I as fighting countries realized that helium was an excellent way to lift military airships. Unlike hydrogen-filled airships, those carrying helium didn't explode when the enemy shot them down. And helium's properties, particularly its ability to be liquid at close to absolute zero (minus 459.67 degrees Fahrenheit, or minus 273.15 degrees Celsius), meant it could cool machine parts like engines and magnets better than any other element could.

Today, helium is an essential cooling component in nuclear reactors, rockets and medical diagnostic equipment such as MRI machines. The gas keeps fiber optics, superconductors, quantum computers and semiconductors cool, but skyrocketing demand has pushed supply chains to their limit, resulting in a global shortage that has persisted for more than a decade. Helium extraction also has a huge carbon footprint — almost equivalent to the U.K.'s per year — because currently, it is exclusively produced together with natural gas.

However, in recent years, pioneering discoveries have led to a pivotal change in scientists' understanding of the geology that helps helium accumulate. Researchers have uncovered reservoirs of primary, "carbon-free" helium — large accumulations of the gas that are highly concentrated and don't contain methane — that could revolutionize the industry.

This new understanding has fueled exploration projects in a handful of regions around the world. From Yellowstone to Greenland to the East African Rift, a helium "rush" is starting to address shortages and helium's enormous carbon footprint.

"It's a new industry," Thomas Abraham-James, co-founder and CEO of the exploration company Pulsar Helium, told Live Science.

Perfect seals, imperfect yields

After World War I, helium discoveries multiplied and the U.S. emerged as the world's leading producer. Gas wells with helium levels of 0.3% and above were tapped to fuel a growing number of industries and form a stockpile, the Federal Helium Reserve in Amarillo, Texas. (The stockpile was sold in 2024 to the industrial gas firm Messer.)

But helium has been produced only as a minor byproduct, usually found in tiny amounts mixed in with natural gases such as methane.

That's because gases like methane and carbon dioxide (CO₂) are required to transport helium from the middle part of Earth's crust to shallower regions, Chris Ballentine, a professor of geochemistry at the University of Oxford, told Live Science in an email.

Helium forms in the top 16 miles (25 kilometers) of Earth's crust when uranium and thorium break down into other radioactive elements, emitting alpha particles, or helium nuclei, in the process. These helium nuclei gain two electrons from atoms in their environment to form helium atoms, which then migrate and gradually collect in groundwater down to 10 miles (16 km) below Earth's surface, Ballentine said.

But for helium atoms to form a gas, they need to reach their "bubble point" — the concentration of a dissolved gas in a liquid required to make buoyant gas bubbles. Helium in groundwater rarely accumulates in sufficient quantities to reach this bubble point, Ballentine said. So other, more ubiquitous gases, like methane and CO₂, are typically needed to entrain helium and form the bubbles that rise toward geological traps, Ballentine said.

These geological traps are often natural gas fields. But helium also gets trapped, because natural gas fields often have strong seals, research suggests.

Natural gas reservoirs form beneath layers of fine-grained "cap" rocks "and minerals. In Australia's Amadeus Basin, for example, helium and natural gas are locked beneath a thick layer of salt, which is a "perfect" seal, Jon Gluyas, a professor of geoenergy, carbon capture and storage at Durham University in the U.K., told Live Science.

Most places, however, don't have a perfectly sealed reservoir to trap helium underground, so the gas escapes into the atmosphere. "All systems are leaky," Gluyas said. As a result, most regions with helium-producing rocks emit some helium, so "if you went out with a sensitive enough instrument, you would find it," he said.

Helium's association with natural gas means producers can extract both at the same time, but this approach has major drawbacks. For one, the helium industry currently has an indirect carbon footprint of about 350 million tons (320 million metric tons) per year, which is bigger than all but 20 countries' carbon footprints.

A second downside is that countries with natural gas deposits and the companies drilling these deposits control the world's supply of helium. The U.S. used to dominate global helium production, but Qatar took the lead in 2022. Depending on other countries for helium introduces risks related to regional geopolitics, Gluyas said. Algeria and Russia are also leading helium producers, raising similar concerns.

But the biggest problem with helium extraction is that natural gas contains minuscule amounts of helium. In the U.S., the lower profitable limit to separate helium from natural gas is 0.3%, but some countries with different production and transport methods have much smaller thresholds. For example, Algeria's Hassi R'Mel gas field contains 0.19% helium and Qatar's North Dome deposits contain 0.04% helium, yet both countries extract helium at these locations.

Groundbreaking discoveries

Data show that only about 1 in 6 natural gas reservoirs in the U.S. has helium levels higher than 0.3% and that levels above 7% are extremely rare, meaning economically useful helium concentrations are the exception in the U.S.

The trend is similar in other countries. So when scientists found a nitrogen gas reservoir that contained up to 10.4% helium in Tanzania in 2016, they were flabbergasted. The gas was bubbling out of the ground in the Rukwa Rift Basin, which is located on a divergent plate boundary called the East African Rift.

Crucially, the Rukwa Rift Basin doesn't have reservoirs of natural gas or other hydrocarbons. This was the first major confirmed discovery of a hydrocarbon-free helium reservoir, and it sparked an ongoing worldwide hunt for other such reservoirs.

Ballentine, Gluyas and Abraham-James were part of the team that made the discovery in Tanzania. "The approach we adopted essentially is similar to that which any explorer would use for petroleum — we went looking for seeps," Gluyas said. Helium seeps are "all over the show," he said, but locating them is trickier than finding petroleum because helium is odorless, colorless and usually at low concentrations.

From then on, researchers and exploration companies tried to determine whether known helium seeps led to hydrocarbon-free reservoirs with high concentrations of the noble gas. In 2021, for example, the exploration company Pulsar Helium acquired land near Babbitt, Minnesota, where a company searching for nickel had previously found gas with high helium concentrations. Pulsar Helium drilled a well down to 2,200 feet (670 meters) in early 2024 and found a huge gas reservoir with helium concentrations up to 14.5% — the highest the industry has ever seen in North America.

The other gases in the reservoir were nitrogen and CO₂. But the site contains no natural gas, so it counts as a primary helium accumulation, Pulsar Helium representatives say. CO₂ concentrations were above 70%, but the company views this as an opportunity rather than a problem, because the gas is pure enough to be used for carbonated beverages, water treatment, food preservation and medicine.

Unique geology

The more reservoirs of hydrocarbon-free helium that researchers and companies find, the more information they learn about the geology that forms these reservoirs. Shortly after the discovery in Tanzania, geologists identified five key conditions needed to form accumulations of helium without natural gas.

First, the region must have helium-producing rocks miles below the surface. The best helium source rocks are ones that contain uranium or thorium, and are usually composed entirely of crystallized minerals, Gluyas, Ballentine and their colleagues wrote in a 2024 article for the Energy Geoscience Conference Series.

That's because crystalline rocks solidify from magma cooling extremely slowly underground. This unhurried process concentrates radioactive uranium and thorium, which are unstable elements in mineral structures and therefore among the last to be incorporated. Granite is one of the best helium sources, Ballentine said.

Ideally, the rocks should also be hundreds of millions to billions of years old, because the radioactive versions of uranium and thorium that decay into alpha particles have half lives of 4.5 billion and 14 billion years, respectively, according to the article. That means it would take that long for half of a sample of those elements to decay into helium. As a result, it takes millions of years to accumulate enough helium to fill a large reservoir.

The second criterion needed to form a hydrocarbon-free helium reservoir is a heat source around the rocks.

Normally, helium is frozen in a mineral's lattice structure because that structure is "blocked" — or doesn't exchange molecules with its surroundings. For minerals to release that helium, they must exceed their "closure temperature" — the temperature at which the lattice becomes unblocked. That temperature varies, but can be above about 160 F (70 C) for one of the most common helium-containing minerals.

Often the source of heat is volcanism or geothermal heat, so hydrocarbon-free helium reservoirs typically occur in regions where there is, or once was, volcanism. Beneath the Rukwa Rift Basin, for example, eastern Africa is pulling away from the rest of the continent, causing magma to rise to the surface. Similarly, Pulsar Helium's exploration site in Minnesota sits on an ancient tear in North America's crust called the Midcontinent Rift System. This rift system started and then failed about 1.1 billion years ago, producing intense volcanic activity during the roughly 100,000 years it was active.

The third condition to form a hydrocarbon-free helium reservoir is the presence of nitrogen in groundwater, because nitrogen bubbles can transport helium upward through the crust in the same way methane and CO₂ bubbles can, while completely removing greenhouse gases.

The fourth condition is the need for relatively airtight "cap rocks" that sit near the surface above the helium-forming rocks in Earth's crust. That's because when nitrogen reaches its bubble point, it captures helium atoms and taxies them until the gases either escape into the atmosphere or are trapped. But for that to happen, the cap rocks must form an impermeable seal.

The fifth and final condition is that these cap rocks must sit atop fractured, porous "reservoir" rocks that can store gas, Ballentine, Gluyas and their colleagues wrote in the 2024 article.

The growth rate of a reservoir depends on the rates at which gases enter from below and escape via cracks in the seal, the researchers wrote. The more helium that enters a reservoir and the less that escapes through the seal, the bigger an accumulation can get. In other words, reservoirs ideally have porous or highly fragmented rocks down below, and nonporous, intact rocks above.

Helium reservoirs with impermeable seals can hold gases for long periods of geological time. For example, the link between the reservoir in Minnesota and the Midcontinent Rift System suggests helium has been building up there for 1.1 billion years.

Appraisal and development

Pulsar Helium recently announced that it will start engineering work for a helium production plant at its site in Minnesota, signaling that hydrocarbon-free, U.S.-produced helium could reach the market in just a few years.

Early this year, the company more than doubled the depth of its first well to reach the bottom of the reservoir and drilled a second well down to 5,638 feet (1,718 m). Tests over the summer showed high flow rates to the surface and stable helium concentrations of up to 8% at both wells, "providing a robust foundation for future development," Abraham-James said in a September statement. Since then, Pulsar Helium has drilled an additional well down to 3,507 feet (1,069 m) and opened two more wells.

Pulsar Helium is also progressing with a helium project in East Greenland — the first helium discovery on the island, Abraham-James told Live Science.

"What we've learned in Minnesota and elsewhere, we then applied it to Greenland and we found the helium there," he said. "Like Minnesota, its helium is not associated with hydrocarbons. We conducted a seismic survey last year, and that went some ways to mapping the reservoir."

Located about 3 miles (5 km) from the coastal settlement of Ittoqqortoormiit, the site in East Greenland looks promising for helium and geothermal energy production, which could limit the settlement's dependency on fossil fuels, Abraham-James said. Mapping in 2024 revealed a zone of crust with temperatures reaching 266 F (130 C), as well as a fractured reservoir that researchers linked to gas emissions at the surface containing up to 0.8% helium, according to a statement from Pulsar Helium.

Any helium produced in East Greenland would likely go to the local community, Abraham-James said. Similarly, helium produced in Minnesota would be sold inside the U.S. to supply MRI scanners, semiconductor fabrication and space launches, he said.

Helium shortages in the U.S. have eased somewhat since early 2024, partly thanks to additional supplies from natural gas fields, said Nicholas Fitzkee, a professor of chemistry at Mississippi State University. "But having a larger domestic supply would be valuable, because it could insulate the U.S. from geopolitical instabilities that have contributed to past helium shortages," Fitzkee told Live Science in an email.

Halfway across the globe, prospecting in Tanzania is also ongoing, with two exploration companies currently reporting helium levels of 5.5% and 2.46% at different ends of the Rukwa Rift Basin. Called Helium One Global and Noble Helium, these companies are still in the early phases of exploration, Gluyas said.

"Both Helium One and Noble Helium have successfully shown elevated concentrations of helium in the wells they've drilled," he said. But these wells don't provide a clear picture of the reservoir yet, and the reservoir may not end up fulfilling researchers' expectations.

"They could speculate, based upon the seismic information, what the geometry of the accumulation might be, but they haven't yet drilled sufficient wells to say, 'It really is that,'" Gluyas said.

Beyond Tanzania, there may be opportunities for helium exploration in India's Bakreswar-Tantloi geothermal area, which is located in the east of the country and straddles the states of West Bengal and Jharkhand. The Bakreswar-Tantloi area sits on ancient granitic rocks that are rich in uranium and therefore produce helium. The region also has a fault system and a high heat gradient as a result of ongoing tectonic activity along the Son-Narmada-Tapti rift zone, research suggests.

Closer to home, researchers are analyzing the conditions for potential helium accumulations beneath and around Yellowstone National Park. Yellowstone is rooted in the Wyoming Craton, a prehistoric region of crust and upper mantle that contains 3.5 billion-year-old rocks known to produce huge quantities of helium. Thanks to Yellowstone's countless geothermal features and volcanic structures, helium may be accumulating in reservoirs beneath or peripheral to the park, although it's more likely that the gas is circulating and escaping into the atmosphere through a complex system of natural pipes.

"What's happened over the millions or hundreds of millions of years in the area in which Yellowstone occurs is that helium has been building up, and now in the last [roughly] 5 million years, the supervolcano beneath is flushing it out," Gluyas explained.

That means the chances of extracting this helium are remote, not least because of the scorching temperatures of up to 275 F (135 C) that drillers would encounter belowground. "Will your drilling equipment survive? Almost certainly not," Gluyas said.

But beyond Yellowstone, it's important for existing hydrocarbon-free helium projects to complete their evaluations and start selling the gas as soon as possible, Gluyas said. "There is a huge need for helium," he said.

However, Fitzkee sees another way forward — rolling out technologies that can recycle or lower our helium consumption. That could be through installing helium recovery systems or engineering room-temperature alternatives to current helium-hungry technologies, he added.

Topping up our helium supply is a good stopgap solution, but not a permanent fix, he argued.

"Ultimately, we cannot mine our way out of future helium shortages," he said. "Helium is non-renewable, and we have no easy way to make more at scale."

Imagine driving down a highway in the summer. The windows are down, the music is loud, and the wind is whipping through your hair. Now picture your car's windshield. You might expect to see a handful of splats from unfortunate bugs. But 30 years ago, there would have been significantly more buggy skid marks plastered on the front of your vehicle.

"When I was a kid, you could go out driving in the summer, and you would come home and your car windshield was covered in bugs," said Cheryl Schultz, an ecologist at Washington State University. "Now, you can go across many areas at the same time of year and your windshield is clean."

This phenomenon, called the "windshield test," is indicative of a larger, very worrying trend: Insects, particularly the flying ones that pollinate many crops, are in steep decline. This nosedive is disrupting ecosystems around the world, and could jeopardize the global food supply. But tracking the decrease of insect populations over the past three decades has proved tricky — and stopping the decline may be even harder.

However, researchers are working quickly to find ways to stem the tide and even reverse the trend. Key to that is a collaborative approach that includes local and federal conservation efforts, new pollinator habitats, and a reduction in pesticide use.

The age of the "insect apocalypse"

Both the total number of insects and the number of insect species have been declining for decades in pretty much every place scientists have looked — prompting researchers to dub it "the insect apocalypse." Global bee biodiversity is down 25% compared with pre-1995 numbers, according to research published in 2021. A sweeping 2025 study showed that butterfly abundance across the U.S. fell by 22% over the past two decades. And a study in Germany found a whopping 76% loss of flying insects in some of the country's forested areas over 27 years.

"It's a worrisome thing," Scott Black, executive director of the nonprofit Xerces Society for Invertebrate Conservation, told Live Science.

By and large, experts know why insects are becoming scarcer. The first factor is climate change. As the planet warms, key host plants for insects start to bloom earlier each year. This can cause a mismatch in life cycles for certain species, putting many newly hatched or metamorphosed bugs out of sync with their food sources. And extreme heat, reduced snowpack, severe storms and megadroughts can chip away at previously robust insect numbers. Many populations simply can't keep up. Meanwhile, milder winters can benefit a few adaptable pest species, which may outcompete sensitive insects and wreak ecological and agricultural havoc in some regions.

The second driver is habitat loss — the inexorable creep of urbanization, deforestation and sterile suburban lawns, which host fewer and less-diverse ranges of insects. As humans encroach on insect habitats, insects like ground-dwelling bees are left without space to build nests, rear young and overwinter, leading to population declines.

Finally, there are pesticides. For instance, neonicotinoids (often labeled as the active ingredients acetamiprid, clothianidin, dinotefuran, imidacloprid and thiamethoxam), have been identified as a major threat to wild bees, and they're still used in the U.S. and some other industrialized countries, including parts of Canada and Australia. Other pesticides, like the common weed killer glyphosate, have been shown to weaken bees' ability to regulate hive temperature, leaving them vulnerable to plunging winter temperatures.

"It's really extremely rapid environmental changes that we're seeing," Roel van Klink, a researcher at the German Center for Integrative Biodiversity Research, told Live Science. "Those species that were adapted to the conditions that we had maybe 50 or 100 years ago are not adapted to the conditions now anymore. And so they go down."

Collecting data on the scale and scope of these declines has been challenging, however. For one thing, some insects are easier to find than others. Flying insects like beetles and dragonflies are much more mobile, and therefore easier to spot, than earthbound bugs like earwigs and ants. Likewise, charismatic insects like bees and butterflies tend to have more historical records of their numbers and are usually easier to identify.

But there's another reason these insects' declines have gotten more scientific attention: They are extremely important for global food security.

The importance of diverse pollinators

Disappearing insects are bad news for the global food system. As the world's population continues to grow, the stress that insect declines — and dropping pollinator numbers, in particular — put on the food system could lead to an agricultural economic collapse, as well as increased food scarcity.

"Preventing further declines is no longer enough," Francesca Mancini, an ecological modeler at the UK Centre for Ecology & Hydrology, told Live Science. "We need to restore insect biodiversity to past levels."

In the U.K. alone, insect pollinators provide an estimated $1 billion in economic value each year, Mancini said. For the U.S., it's in the ballpark of $34 billion.

Worldwide, three-quarters of the crops we eat — and just over one-third of total crop yields — depend on pollination by insects. The degree to which these crops rely on pollinators falls along a spectrum. Some, like soybeans, would be much less productive without insect pollination. Others would cease to exist. "Coffee and chocolate are actually 100% dependent on pollination by insects," van Klink said.

A lot of that pollination work is done by managed European honeybees (Apis mellifera), which beekeepers around the world diligently maintain, transport and unleash upon fields across the globe each year. But to flourish, many crops need more than just honeybees.

For example, fruits native to North America, like blueberries and tomatoes (which is technically a fruit), are more effectively pollinated by native bumblebees, such as Bombus fraternus. That's because bumblebees can perform what's known as "buzz pollination," where they land on a flower and vibrate rapidly to release even the most deeply held pollen grains. Cacao trees (Theobroma cacao) — the source of the cocoa beans used to make chocolate — are entirely pollinated by chocolate midges. And cotton yields would plummet by up to 50% without butterfly pollinators.

Some staple crops, like soybeans, can make it without insects. However, research has shown that soybean fields visited by pollinators have significantly higher yields.

Then, there are crops like alfalfa (Medicago sativa). This legume isn't widely consumed by humans, but it is a staple for livestock — particularly dairy and beef cattle. Like blueberries and tomatoes, alfalfa depends on insect pollinators to thrive. However, honeybees will only pollinate it reluctantly; given the choice, they'd rather buzz around plants with flowers that are easier for them to access. But wild bees, particularly the alfalfa leaf-cutting bee (Megachile rotundata), are extremely effective alfalfa pollinators.

A recent study found that alfalfa fields visited by a mix of honeybees, wild bees and other pollinators, like wasps and butterflies, produced significantly more and larger seeds than fields visited by honeybees alone. This higher yield translates to more food for cattle — and thus more milk, cheeseburgers and steaks for us.

Glimmers of hope

Of course, restoring insect abundance and biodiversity is no easy task, especially in the face of an all-encompassing threat like global climate change. Experts told Live Science that coordinated federal regulations aimed at slowing climate change, reducing industrial pesticide use, and preventing the destruction of wild spaces are essential for protecting insects. But there are also actions people can take at the local and personal level that can have a positive impact.

Although the current U.S. administration's cuts to federal science programs and green energy have dealt a harsh blow to progress on these fronts, many experts still see reasons for optimism.

"As much as the overall picture is overwhelming, there's lots of places for hope," Schultz told Live Science. In a detailed report about the state of U.S. butterflies written this year in collaboration with the Xerces Society, Schultz highlighted a number of "success stories" — species that bucked the trend and increased in abundance thanks to years of focused work at both the federal and local levels.

Chief among them is the Fender's blue (Icaricia icarioides fenderi), a tiny azure butterfly native to Oregon. In 2000, the U.S. Fish and Wildlife Service listed it as endangered. In 2023, it became the second-ever insect to be downlisted to "threatened."

And the benefits of conservation efforts for one species had knock-on effects: Of the 342 butterfly species and subspecies analyzed in the report, 65 others had increased in number, and most were not on the endangered species list. This suggests that protections to conserve one insect could benefit others as well.

Increasing healthy habitat

One of the best ways to help butterflies and other pollinators is to create more habitat for them. Unlike grizzly bears or elk, these insects don't need large stretches of unbroken wilderness. Even something as small as a backyard butterfly garden or a flower-filled window box can go a long way, Wendy Leuenberger, an ecologist at Michigan State University, told Live Science.

One study in the Pacific Northwest found that converting a 5,400-square-foot (500 square meter) plot of land — roughly half the size of the average American lawn — into an insect-friendly habitat full of native or wild plants can increase pollinator species' richness and abundance by about 90%. However, that effect was fairly localized, and it dissipated when these patches were placed in plots of more than 150,000 square meters (37 acres) — about the size of seven or eight blocks in Chicago.

Some pollinators, like hoverflies (Syrphidae spp.) and certain types of bees, can cover miles in search of flowering plants. But others, including many butterflies, tend to stay closer to home — within a 650-foot 200 meter radius for more delicate species. This suggests that plots of native or wild flora are most effective at bolstering our food supply when interspersed within larger agricultural fields.

"I would say it's the closer, the better for your crops," Andy Grinstead, a conservation manager at Pollinator Partnership, told Live Science.

In agricultural communities, experts like Grinstead recommend planting "buffer strips" of native vegetation near (or, if possible, in between) crops. He also suggests planting hedgerows of woody, flowering plants around fields to act as both pollinator habitat and wind protection. But you don't have to be a farmer to support pollinators. Folks living within a few miles of farms can plant "bee lawns," which are filled with low-growing flowering plants like clover, instead of pure turfgrass.

And for those without yards, growing micro-plots of native wildflowers — even just a pot on a rooftop or balcony or hanging from a window — can create green "stepping stones" for bees, hoverflies, migratory butterflies and beetles passing through urban areas.

"Pollinator-friendly practices are valuable across all landscapes," Grinstead said. "It takes very little space to actually make an impact."

Reducing pesticide use on an industrial scale can also benefit pollinators, Black said.

One way to do this is to adopt an integrated pest management framework. This can mean rotating crops to keep soil healthy; accurately identifying pests before applying pesticides; and carefully spraying in targeted areas (away from blooms) when the wind is low to prevent the pesticides from drifting into the surrounding environment.

But even home gardeners can help reduce pesticides by replacing lawns or ornamental plants with hardier native species, hand-weeding rather than blanket-spraying small plots, and using screens or draining standing water instead of spraying for pests like mosquitoes, Black said. Taken together, these actions can help create havens where pollinators can thrive.

Taking action

Crucially, scientists are still researching the full scope of global insect declines, especially for species that have been historically understudied. This means we need field research to estimate insect numbers, Black said.

Community pollinator counts, whether as part of a formal program or through apps like iNaturalist, are also essential, Leuenberger told Live Science. These data help experts pinpoint which species are most vulnerable and which conservation efforts are most effective.

But with the future of the global food system hanging in the balance, it's important to try to restore these numbers now — not wait till researchers have published comprehensive data on how and where insect numbers are plummeting, Black said. "We don't want to wait until we have everything tucked into a perfect paper before we take action," he said. "We know how to take action."

Quantum computers are coming. And when they arrive, they are going to upend the way we protect sensitive data.

Unlike classical computers, quantum computers harness quantum mechanical effects — like superposition and entanglement — to process and store data in a form beyond the 0s and 1s that are digital bits. These "quantum bits" — or qubits — could open up massive computing power.

That means quantum computers may solve complex problems that have stymied scientists for decades, such as modeling the behavior of subatomic particles or cracking the "traveling salesman" problem, which aims to calculate the shortest trip between a bunch of cities that returns to its original destination. But this massive power also may give hackers the upper hand.

"Like many powerful technologies, you can use [quantum computing] for great good," Rebecca Krauthamer, a technological ethicist and CEO of cybersecurity firm QuSecure, told Live Science. "And you can also use it for malicious purposes."

When usable quantum computers first come online, most people — and even most large organizations — will still rely on classical computers. Cryptographers therefore need to come up with ways to protect data from powerful quantum computers, using programs that can run on a regular laptop.

That's where the field of post-quantum cryptography comes in. Several groups of scientists are racing to develop cryptographic algorithms that can evade hacking by quantum computers before they are rolled out. Some of these cryptographic algorithms rely on newly developed equations, while others are turning to centuries-old ones. But all have one thing in common: They can't be easily cracked by algorithms that run on a quantum computer.

The foundations of cryptography

Cryptography dates back thousands of years; the earliest known example is a cipher carved into ancient Egyptian stone in 1900 B.C. But the cryptography used by most software systems today relies on public key algorithms. In these systems, the computer uses algorithms — which often involve factoring the product of two large prime numbers — to generate both a public key and a private key. The public key is used to scramble the data, while the private key, which is available only to the sender, can be used to unscramble the data.

To crack such cryptography, hackers and other malefactors often must factor the products of very large prime numbers or try to find the private key by brute force — essentially throwing out guesses and seeing what sticks. This is a hard problem for classical computers because they have to test each guess one after another, which limits how quickly the factors can be identified.

A 100-story skyscraper on a three-story building

Nowadays, classical computers often stitch together multiple encryption algorithms, implemented at different locations, such as a hard disk or the internet.

"You can think of algorithms like building bricks," Britta Hale, a computer scientist at the Naval Postgraduate School, told Live Science (Hale was speaking strictly in her capacity as an expert and not on behalf of the school or any organization.) When the bricks are stacked, each one makes up a small piece of the fortress that keeps out hackers.

But most of this cryptographic infrastructure was built on a foundation developed in the 1990s and early 2000s, when the internet was much less central to our lives and quantum computers were mainly thought experiments. "It's like a foundation for a three-story building, and then we built a 100-story skyscraper on it," Michele Mosca, co-founder and CEO of cybersecurity company evolutionQ, told Live Science. "And we're kind of praying it's OK."

It might take a classical computer thousands or even billions of years to crack a really hard prime factorization algorithm, but a powerful quantum computer can often solve the same equation in a few hours. That's because a quantum computer can run many calculations simultaneously by exploiting quantum superposition, in which qubits can exist in multiple states at once. In 1994, American mathematician Peter Shor showed that quantum computers can efficiently run algorithms that will quickly solve prime-number factoring problems. As a result, quantum computers could, in theory, tear down the cryptographic fortresses we currently use to protect our data.

Post-quantum cryptography aims to replace obsolete building blocks with less-hackable bricks, piece by piece. And the first step is to find the right math problems to use. In some cases, that means returning to equations that have been around for centuries.

Currently, the National Institute of Standards and Technology (NIST) is looking at four problems as potential foundations for post-quantum cryptography. Three belong to a mathematical family known as structured lattices. These problems ask questions about the vectors — mathematical terms that describe direction and magnitude between interconnected nodes — like the connection points in a spiderweb, Mosca said. These lattices can theoretically have an infinite number of nodes and exist in multiple dimensions.

Experts believe lattice problems will be hard for a quantum computer to crack because, unlike some other cryptographic algorithms, lattice problems don't rely on factoring massive numbers.

Instead, they use the vectors between nodes to create a key and encrypt the data. Solving these problems may involve, for example, calculating the shortest vector in the lattice, or trying to determine which vectors are closest to one another. If you have the key — often a "good" starting vector — these problems may be relatively easy. But without that key, they are devilishly hard. That's because no one has devised an algorithm, like Shor's algorithm, that can efficiently solve these problems using quantum computing architecture.

The fourth problem that NIST is considering belongs to a group called hash functions. Hash functions work by taking the virtual key for unlocking a specific point on a data table, scrambling that key and compressing it into a shorter code. This type of algorithm is already a cornerstone of modern cybersecurity, so in theory, it should be more straightforward to upgrade classical computers to a quantum-proof version compared with other post-quantum cryptographic schemes, Mosca said. And similarly to structured lattices, they can't easily be solved by brute force alone; you need some clue as to what's going on inside the "black box" key generator to figure them out within the age of the universe.

But these four problems don't cover all of the potentially quantum-safe algorithms in existence. For example, the European Commission is looking at an error-correcting code known as the McEliece cryptosystem. Developed more than 40 years ago by American engineer Robert McEliece, this system uses random number generation to create a public and private key, as well as an encryption algorithm. The recipient of the private key uses a fixed cipher to decrypt the data.

McEliece encryption is largely considered both faster and more secure than the most commonly used public-key cryptosystem, called Rivest-Shamir-Adleman. As with a hash function, would-be hackers need some insight into its black-box encryption to solve it. On the plus side, experts consider this system very safe; on the downside, even the keys to unscramble the data must be processed using extremely large, cumbersome matrices, requiring a lot of energy to run.

A similar error-correcting code, known as Hamming Quasi-Cyclic (HQC), was recently selected by NIST as a backup to its primary candidates. Its primary advantage over the classic McEliece system is that it utilizes smaller key and ciphertext sizes.

Another type of algorithm that sometimes comes up in conversations about post-quantum cryptography is the elliptic curve, Bharat Rawal, a computer and data scientist at Capitol Technology University in Maryland, told Live Science. These problems go back at least to ancient Greece. Elliptic curve cryptography exploits basic algebra — calculating the points on a curved line — to encrypt keys. Some experts believe a new elliptic curve algorithm could evade hacking by a quantum computer. However, others argue that a hacker could hypothetically use Shor's algorithm on a quantum computer to break most known elliptic curve algorithms, making them a less-secure option.

No silver bullet

In the race to find quantum-safe cryptographic equations, there won't be a silver bullet or a one-size-fits-all solution. For example, there's always a trade-off in processing power; it wouldn't make much sense to use complex, power-hungry algorithms to secure low-priority data when a simpler system might be perfectly adequate.

"It's not like one algorithm [combination] will be the way to go; it depends on what they're protecting," Hale said.

In fact, it's valuable for organizations that use classical computers to have more than one algorithm that can protect their data from quantum threats. That way, "if one is proven to be vulnerable, you can easily switch to one that was not proven vulnerable," Krauthamer said. Krauthamer's team is currently working with the U.S. Army to improve the organization's ability to seamlessly switch between quantum-safe algorithms — a feature known as cryptographic agility.

Even though useful (or "cryptographically relevant") quantum computers are still several years away, it is vital to start preparing for them now, experts said. "It can take many years to upgrade existing systems to be ready for post-quantum cryptography," Douglas Van Bossuyt, a systems engineer at the Naval Postgraduate School, told Live Science in an email. (Van Bossuyt was speaking strictly as a subject-matter expert and not on behalf of the Naval Postgraduate School, the Navy or the Department of Defense.) Some systems are tough to upgrade from a coding standpoint. And some, such as those aboard military craft, can be difficult — or even impossible — for scientists and engineers to access physically.

Other experts agree that post-quantum cryptography is a pressing issue. "There's also the chance that, again, because quantum computers are so powerful, we won't actually know when an organization gets access to such a powerful machine," Krauthamer said.

There's also the threat of "harvest-now, decrypt-later" attacks. Malicious actors can scoop up sensitive encrypted data and save it until they have access to a quantum computer that's capable of cracking the encryption. These types of attacks can have a wide range of targets, including bank accounts, personal health information and national security databases. The sooner we can protect such data from quantum computers, the better, Van Bossuyt said.

And as with any cybersecurity approach, post-quantum cryptography won't represent an end point. The arms race between hackers and security professionals will continue to evolve well into the future, in ways that we can only begin to predict. It may mean developing encryption algorithms that run on a quantum computer as opposed to a classical one or finding ways to thwart quantum artificial intelligence, Rawal said.

"The world needs to keep working on this because if these [post-quantum equations] are broken, we don't want to wait 20 years to come up with the replacement," Mosca said.

On Dec. 31, 2019, the first reports emerged about a mysterious pneumonia of unknown cause circulating in China. On March 11, 2020, the World Health Organization declared COVID-19 a pandemic.

On March 16, the first COVID-19 vaccine entered clinical trials.

And by Dec. 14, members of the American public were getting the first doses of COVID-19 vaccines outside of trials.

The first coronavirus vaccines were paradigm-shifting because they went from conceptualization to mass production in mere months. But they were also unique because they used a new way to stimulate the immune system — one that had been thoroughly studied for decades in order to be ready for deployment at this crucial moment.

The key to these vaccines was messenger RNA (mRNA), DNA's less-famous cousin. The power of the mRNA platform is that vaccines can be produced exceptionally quickly once a pathogen's genetics have been analyzed; conventional vaccine manufacturing takes months or years whereas mRNA vaccines can be made in mere weeks. So while it was once the subject of high school biology classes and niche pockets of biomedical science, mRNA was suddenly thrust into the public eye — and once there, it inspired relentless misinformation and controversy.

While mRNA-based COVID-19 vaccines are the best known application of the molecule, researchers around the world have found additional uses for mRNA technology beyond the realm of vaccines. They're exploring its use for groundbreaking treatments for cancer and autoimmune disease, as well as for gene-editing therapies for genetic disorders. But that promise may be unrealized in the United States, where the federal government has declared war against this promising technology.

This new stance runs counter to the Trump administration's prior embrace of mRNA vaccines.

"We do really have to give President Trump credit for introducing the mRNA platform to the world through his leadership in Operation Warp Speed," said Jeff Coller, the Bloomberg distinguished professor of RNA biology and therapeutics at Johns Hopkins University. "The president should be taking a victory lap." But instead, the second Trump administration is actively dismantling this legacy, Coller told Live Science.

Vaccine skeptic Robert F. Kennedy Jr. now heads the Department of Health and Human Services (HHS), and opponents of both conventional and mRNA-based vaccines hold seats on the country's most influential vaccine advisory committee. Since Trump's inauguration, federal scientists have faced mass layoffs, funding freezes, and memos warning them to disclose their involvement in research areas the administration has targeted, including mRNA vaccines.

These actions had an immediate chilling effect on mRNA research and development in the U.S., Coller told Live Science. And then, in August, HHS canceled nearly half a billion dollars of investment into mRNA vaccine development.

"I was shocked to see this, frankly," said Jordan Green, head of the Biomaterials and Drug Delivery Laboratory at Johns Hopkins, whose lab is developing both mRNA therapies and delivery systems to get the molecule into the body.

The cuts are sending ripples through biotech, making stakeholders question whether it's safe to set up shop in the U.S., or whether their mRNA investments would be better spent abroad. "It's just a shame because it's an unforced error; there's no reason," Green said.

For now, HHS appears to be retreating primarily from mRNA vaccines; it noted "other uses of mRNA technology" would not be affected by the cuts. But "the industry doesn't trust that," Coller told Live Science.

According to Grant Witness, a project tracking scientific grants under the Trump administration, mRNA research unrelated to vaccines has already been hit by grant terminations and funding freezes. So even if it's not being explicitly targeted, it's not necessarily being preserved. The project's database shows that the National Institutes of Health (NIH), part of HHS, terminated grants for projects developing mRNA-based treatments for cancer, Alzheimer's, pulmonary arterial hypertension and HIV, as well as grants for basic research about how mRNA works in healthy and diseased cells.

Here's what the United States stands to lose if the federal government broadly divests of mRNA medicines after spending decades readying the technology for prime time.

Cancer treatment

On a molecular level, mRNA is a close cousin of DNA — and human cells are stuffed with it. These ubiquitous "messengers" copy instructions from DNA and relay them to other locations in the cell — namely, to protein-construction sites where the complex molecules that do most of the work in cells get made. mRNA also performs other key jobs in cells, such as helping control which genes are switched on and to what degree.

For decades prior to the pandemic, dedicated scientists pored over mRNA, learning how this class of molecules works in the body and how it might be leveraged to heal the sick and guard against disease.

"I started doing this when I was like 21, back in the '90s, back when about 10 people on Earth knew what mRNA were," Coller said. Findings from Coller's lab later helped inform the development of Spikevax, the COVID-19 vaccine made by Moderna. But while COVID-19 vaccines are the best-known application of mRNA to date, they're far from the first.

The first mRNA therapeutics company was founded in 1997, and rather than targeting infectious disease, it had its sights set on cancer treatment. Its approach ultimately hasn't panned out in human trials, but in the meantime, other approaches to mRNA-based cancer treatment have gained traction.

Cancer vaccines are a standout example, but in this context, the term "vaccine" is "a bit of a misnomer," said Dr. Vinod Balachandran, a pancreatic cancer surgeon-scientist and director of The Olayan Center for Cancer Vaccines (OCCV) at Memorial Sloan Kettering Cancer Center. Rather than being given preventatively, like a COVID vaccine, cancer vaccines are "given to patients as a treatment; it is a therapy," he said.

These therapies are similar to conventional vaccines in that they train the immune system to recognize antigens, which are substances that act as "red flags" for foreign invaders, toxins or diseased cells. These include, for example, the spike protein of SARS-CoV-2 — the virus that causes COVID-19 — and certain molecules on cancer cells. Balachandran and colleagues have focused on pancreatic cancer, which has a five-year relative survival rate of only 13% — meaning people newly diagnosed with pancreatic cancer are about 13% as likely to survive the next five years compared with the general population.

The researchers have found that, in the rare long-term survivors of pancreatic cancer, the immune system can recognize the cancer and fight off its recurrence. The team hoped to re-create that immune recognition in other patients by analyzing the genetics of their tumors to see what unique antigens they express. They then create customized vaccines that target those molecules.

"We felt at the time [when we started this work] — this is back in 2017 — that the best technology for rapid custom cancer vaccination for patients was to use RNA," Balachandran said. Once you know a given patient's cancer genetics, a personalized mRNA vaccine that targets multiple antigens can be crafted in a matter of weeks. Conventional vaccines that require the antigens to be grown in the lab and purified would take many months to make.

"For cancer vaccination, speed is essential," Balachandran said. "These are patients who are facing deadly cancers, who require rapid treatment. So we do not have the luxury of waiting."

So far, the team has seen some success. In an early-stage trial, they treated 16 patients who had undergone surgery for pancreatic ductal adenocarcinoma (PDAC), the most common form of pancreatic cancer, which has a five-year survival rate of about 10% to 12%, though they fare a bit better when their cancer can be surgically removed.

Within nine weeks of surgery, Balachandran said, each patient's tumor tissue was analyzed and a personalized vaccine was crafted and delivered alongside post-surgery cancer treatments, like chemotherapy. Half of the patients responded to their vaccine, producing immune cells that have persisted for nearly four years so far, and estimates suggest they could last an average of seven years, "with some lasting even beyond a decade," Balachandran said.

The vaccine responders had a significantly lower risk of recurrence in the following three years than those who didn't respond, with six showing no signs of recurrence in that time frame. A mid-stage trial is now testing the vaccine in about 260 people to see how well it delays or prevents recurrence compared with standard treatment.

"Anything we can do to improve outcomes for these patients who really need help I think will be transformative for the field, for them, for their families — it will mean a lot," Balachandran said. "All other immune therapies and all other therapies have largely failed."

In this initial trial, vaccine production took over two months, in part because samples had to be shipped to overseas collaborators at BioNTech. "We are confident this can happen much faster" in the future, perhaps within a month, Balachandran said.

Other scientists are working on off-the-shelf cancer vaccines that can help bridge the gap until a patient gets a personalized one. These vaccines use a mix of mRNA molecules to stir up a generic, first-line immune defense against cancer. In mice with different kinds of solid tumors, researchers led by Dr. Elias Sayour, a pediatric oncologist at the University of Florida, demonstrated that such a vaccine triggers anti-cancer responses by itself. And when used in combination with another cancer treatment, the mRNA mix can boost the effects of that therapy, Sayour's team has found.

That team has now moved on to human patients, with a two-pronged approach: an off-the-shelf cancer vaccine, followed by a personalized one. They have also run trials of personalized vaccines for the deadly brain cancer glioblastoma, finding that the shots mount a strong, targeted immune response against these tumors that are usually difficult for the immune system to "see." Meanwhile, at another lab, scientists are testing a personalized vaccine in late-stage trials for the skin cancer melanoma and non-small cell lung cancers.

Pancreatic cancer has historically been an elusive target for the immune system because, compared to cancers like melanoma, its cells carry relatively few antigens, Balachandran explained. But pancreatic cancer can be targeted with these vaccines. That bodes well for using mRNA approaches to treat potentially many other cancers, and "I think that's really the most exciting take-home," he said.

Beyond cancer vaccines, Green and colleagues are leveraging mRNA to fight the disease in a different way: by forcing tumors to raise their own red flags to the immune system. Using mRNA packaged inside nanoparticles, the researchers introduce immune-cell genes into cancer cells, prompting tumors to expose their antigens and secrete molecules that call immune cells to the area.

"We can program that tumor cell to now act like an immune cell that helps teach other immune cells what its antigens look like, how to recognize it, how to destroy it," Green told Live Science. In mouse models of breast cancer and melanoma, they combined this approach with an existing immunotherapy and found that it helped shrink and clear tumors from the body while also extending survival.

If shown to work in people, "this could just be off the shelf," he said. It would be one injectable that could work on any patient's solid tumor, he said.

Immune reprogramming

In cancer, scientists are exploring different ways to launch an immune attack against tumor cells, whether by igniting a generalized response or enabling cells to spot specific red flags. But in some diseases, the immune system itself is the culprit — and in those instances, mRNA can help rein in turncoat immune cells.

"Autoimmune diseases — type 1 diabetes, multiple sclerosis, Crohn's, colitis — these can be treated with these mRNA and genetic-based therapies," potentially, Green said. The goal of these treatments would be to "tune the immune system" so it stops attacking healthy tissues.

In type 1 diabetes, for example, the immune system attacks beta cells, which make insulin, leaving the body too little of the hormone to control blood sugar. Green and colleagues are in the early stages of developing an mRNA medicine to reprogram the immune system so it better tolerates beta cells, rather than attacking them. They aim to do this by targeting special immune cells in the liver that promote a tolerant environment in the organ.

The liver is constantly exposed to antigens from food and from microbes in the gut, so immune activity is dialed down to prevent an overreaction. The team's idea is to deliver mRNA that codes for beta-cell proteins to the liver, essentially marking those beta-cell proteins as "safe" to the immune system. This, in turn, can increase the number of regulatory T cells that recognize the proteins as safe; regulatory T cells keep other immune cells in check, and could thus help ward off further attacks on beta cells.

"The problem with autoimmunity is that [the immune system] thinks parts of its own body are foreign," so it's trying to attack them, Green explained. The hope is that "we can use these mRNA medicines to train the immune system so that it sees, 'Oh no, this is fine.'" The work is currently in preclinical stages, as the team runs experiments with cells and lab mice to refine their mRNA nanoparticles.

Green calls this approach of fine-tuning specific cells' activity "genetic surgery." Rather than using a "blunter instrument," like immunosuppressive drugs, to broadly suppress the immune system, the surgery makes a precise change to counter only harmful immune activity. In the long run, the lab is looking to apply this same approach to other autoimmune diseases beyond diabetes, such as multiple sclerosis, in which immune cells target myelin, the insulation surrounding nerve fibers in the brain and spinal cord.

Meanwhile, BioNTech has published data from its own early tests of an mRNA vaccine designed to quell autoimmunity against myelin. In mice, the vaccine expanded populations of immune cells that then kept the myelin-attacking cells in check, suppressing their activity without hobbling the immune system as a whole.

Elsewhere, researchers including Dr. Samira Kiani of the University of Pittsburgh School of Medicine and Kathryn Whitehead of Carnegie Mellon University are using mRNA to counter off-the-rails immune responses, such as the extreme inflammation seen in sepsis. Their approach, described in a preprint, packages mRNA inside a tiny bubble of fat called a lipid nanoparticle (LNP), which then ferries the mRNA into immune cells. From there, the mRNA instructs the cells to make proteins called "zinc finger repressors," which latch onto genes and suppress their activity. In this case, they repressed a key gene involved in the immune signaling that can lead to runaway inflammation.

In experiments with cells in lab dishes and with mice, this type of "epigenetic engineering" showed promise as a potential method for tuning the activity of the immune system. One perk of mRNA is that it degrades quickly in the body, the study authors wrote, so in theory, once the harmful inflammation is subdued, the immune system can get back to defending the body against germs.

Gene editing

mRNA is also being used in tandem with the gene-editing system CRISPR to revolutionize treatments for genetic disease.

Classical CRISPR systems use molecular scissors to snip through DNA strands and enable scientists to tweak specific segments of the molecule. And now, a modified system called base editing can be used to precisely change just one "letter" in DNA's code.

But if you want to use CRISPR for gene therapy, you must first get the scissors, as well as a "guide" molecule that directs them to the right spot — into human cells. And, ideally, the scissors enter only the cells you need to edit, said Giedrius Gasiūnas, a senior researcher in the Life Sciences Center at Vilnius University in Lithuania.

That's where mRNA packaged in nanoparticles could be a game changer.

"This technology could be important for in vivo delivery," meaning the delivery of gene editors directly into the body, said Gasiūnas, who is also chief scientific officer of the biotech company Caszyme.

The first CRISPR-based therapy approved in the U.S. involved editing cells outside the body. The therapy treats two blood disorders by disabling a gene called BCL11A; patients have blood-making stem cells removed from their bone marrow, edited in the lab and then returned to their bodies. But this complex treatment involves a month-long hospital stay, during which the edited stem cells give rise to new blood cells. By comparison, mRNA approaches to gene therapy would be easier to administer and thus more likely to be scalable, Gasiūnas said.

mRNA isn't the only vehicle for getting CRISPR therapies into the body. Some existing gene-editing treatments instead use harmless viruses, such as adeno-associated virus (AAV) vectors, and this same approach is being applied to CRISPR in some emerging therapies. But mRNA packaged inside nanoparticles is emerging as a key player in the field.

This application made headlines in the case of KJ, an infant who became the first person to receive a personalized CRISPR treatment. "The CRISPR technology was introduced as an mRNA," Coller said. "That is the critical feature that was necessary to get this to work."

KJ was born with a severe form of carbamoyl phosphate synthetase 1 (CPS1) deficiency, an inherited disease that causes ammonia to accumulate in the body and affects an estimated 1 in 1.3 million people worldwide. The condition arises from mutations in the CPS1 gene, but in different patients, the gene is broken in different ways. To fix KJ's specific mutation, scientists developed a customized gene therapy consisting of a guide and an mRNA that carried instructions for a base editor into his liver cells, where CPS1 is most active.

"The editor is so large that it cannot easily be accommodated by an AAV vector," said Dr. Kiran Musunuru, co-developer of the therapy and a professor at the University of Pennsylvania's Perelman School of Medicine. Because other potential delivery systems are too early in development, "LNPs are currently the best option," Musunuru told Live Science in an email.

The therapy was ready to administer within six months of KJ's birth. By 9 months old, he had received several doses of the therapy and was thriving; he was hitting milestones he may have never reached otherwise, his medical team reported in May. Had he not received the treatment, he would have needed a liver transplant once he got big enough.

Promising technology at risk

When it comes to the future of mRNA medicines in the U.S., vaccines for infectious disease currently have the bleakest outlook due to political barriers being raised around the technology. Among mRNA tech, vaccines have been the primary target of the federal government's ire, despite experts' warnings that abandoning the shots will leave us vulnerable to pandemics.

"It's foolish to condemn mRNA because people didn't like vaccine mandates or mask mandates or whatever they didn't like," Dr. Seth Berkley, an epidemiologist and senior adviser to the pandemic center at the Brown University School of Public Health, told Live Science. "What we need to do is improve upon them [mRNA vaccines] and understand what the best use of them is," which can only be accomplished through investing in further research.

For now, other applications of mRNA don't appear to be a focus for funding cuts, although some projects have nonetheless been caught in the crosshairs of reduced federal spending. For his part, Musunuru told Live Science he's been in discussions with the Food and Drug Administration regarding ways to get more patients access to personalized gene therapies like KJ's, so that the technology isn't used only in a handful of special cases. When asked about potential funding cuts, he added, "Vaccines and therapies are so different that I do not expect the latter to be affected."

Green, however, anticipates that divestments from vaccines would have unanticipated impacts across the field, because mRNA therapies and vaccines are ultimately composed of very similar genetic molecules and delivery systems.

Regarding nanoparticles, "the vaccine development is … at the pioneering edge of this technology platform," Green said. He argues that cuts to vaccine R&D will inevitably undermine the development of nanoparticles that could be really useful in treating cancer or genetic disease. "It's going to set us back in a very broad way," he said.

The hits to vaccine research also aren't happening in isolation. Myriad federal science grants have been terminated or frozen; NIH agencies may be facing steep funding cuts; and the government shutdown ushered mass layoffs across leading health institutions. Sayour, the oncologist whose lab is developing personalized and universal cancer vaccines using mRNA, said federal grants have historically been critical to his research.

"None of this would be possible without the support of the federal government," Sayour told Live Science. "That is most certainly, 100% true." And that goes for every stage of research, from studies in lab dishes to large clinical trials.

These widespread cuts and more-targeted cuts to mRNA vaccines will not only starve the academic environments where new biotechnologies are typically nurtured, Coller said; they will also drive prospective talent, investors and industry stakeholders in the mRNA space away from the U.S. "In biotech, we're going to see, over the next five to 10 years, a significant 'brain drain,' where other countries build up their infrastructure," he said.

Coller knows this because he is a founding member of the Alliance for mRNA Medicines (AMM), a global organization aimed at advancing and advocating for mRNA medicines. In the spring, AMM surveyed over 100 leaders in the field. They primarily represented U.S.-headquartered pharmaceutical and biotechnology organizations but also included some based in Europe, Canada or Asia. Even prior to the major HHS cuts in August, these stakeholders were feeling squeezed; nearly half reported "already experiencing direct impacts" from federal policy changes tied to mRNA funding.

The cuts caused projects to be scaled back, partnerships to be terminated, and jobs to be lost. About 30% of respondents said that, if faced with further cuts, they might pivot away from mRNA, and 30% said they'd consider moving their operations to other countries. Over 80% agreed that anti-mRNA policies would drive talent away from the U.S. and toward other locales.

Even as scientists contend with these big-picture problems on the national and international stage, skirmishes have also been unfolding at the state level. These battles primarily concern public access to mRNA vaccines but could potentially have broader ripple effects.

A bill still in committee in Iowa would make administering an mRNA vaccine in the state a misdemeanor with a fine of $500 for each shot given. Another bill in committee in South Carolina would prohibit the use of mRNA-based "gene therapies" in the state, but makes an exception for therapies for noninfectious diseases, such as cancer.

It may be that most bills of this ilk never become law and some, like the one raised in South Carolina and two dead bills in Texas, purposely make exceptions for certain uses of mRNA. However, others have been worded in ways that made it unclear whether any mRNA medicine is permissible. If any of such bills ultimately pass at the state level, that could sway where mRNA companies can manufacture products and run clinical trials, Coller said. So, while an exodus of expertise unfolds at a national level, we may see an echo of that at the state level.

The future of mRNA medicine in the U.S. rests on many "ifs," and the worst-case scenario may not come to pass. But in the face of staggering uncertainty, researchers in the field are starting to look for better bets.

"Let's not fool ourselves: mRNA is one of the three most important molecules in the body, with the other two being DNA and protein. It's the intermediary between them," Coller said. "When the federal government sends a message that mRNA-based medicine and research is not wanted, you're basically saying that there's a whole branch of science that is no longer welcome within the U.S."

Around four years ago, now 77-year-old John Gormly went for what was supposed to be a routine blood test. But the results were life-changing.

The test suggested Gormly had colon cancer, which a colonoscopy later confirmed was Stage 2, meaning the cancer had spread through the wall of the colon but not to his lymph nodes.

"I thought [my doctor] was wrong," Gormly, CEO of a construction company near Newport Beach, California, told Live Science. "I go, 'Nah, I don't feel anything.' But there it was. It was real; the colonoscopy showed it."

Gormly was one of the first patients to take a newly approved test called Shield, which its makers say can detect colon cancer from a blood sample. After his diagnosis, Gormly had surgery to remove the tumor and was back to work within 10 days.

An early version of Guardant Health's Shield test has been commercially available since 2022, but it wasn't covered by insurance. However, after approval from the U.S. Food and Drug Administration (FDA) in July 2024, a diagnostic version of Shield was launched commercially and is now covered by Medicare.

Shield is only a blood drop in an ocean of emerging "liquid biopsies."

Scientists have developed blood tests for several cancers, including those of the breast, pancreas and stomach. Some blood tests even detect multiple types of cancer. If these liquid biopsies can be rolled out widely, they could help detect cancer earlier, more easily, or with fewer invasive measures — which, in turn, could lead to earlier detection and fewer cancer deaths.

But many of these tests are still in their early stages. They often detect a lower fraction of cancer cases than gold-standard screening tools like colonoscopies do, meaning they will likely supplement, rather than replace, traditional screening methods. Others may have unacceptable rates of "false positives," meaning a person is initially told they have cancer but diagnostic follow-ups show they do not. This can lead to needless worry or additional invasive tests. These include traditional biopsies, which involve removing tissue samples via needles or surgery. And for some diseases, it's not clear that early diagnosis on a blood test will lead to better outcomes.

However, as these kinks are ironed out, it's likely that blood-based cancer screening will become a normal part of our medical care — one that has the potential to improve cancer outcomes dramatically, experts say.

Simplifying screening

Gormly's doctor recommended a Shield test after noticing that Gormly hadn't had a colonoscopy in a while. He's not alone. Current recommendations suggest that people ages 45 to 75 who are at average risk of colon cancer get a screening, such as a colonoscopy or a stool-based test, every five to 10 years. Yet around 1 in 3 of these people have never been screened.

That's a problem, because colon cancer is the fourth-most-common cancer. Experts have argued that early detection could eliminate 90% of colon cancer deaths. It typically takes around 10 years for early, precancerous growths like polyps to morph into deadly cancer cells, and if these cells are caught early, they can easily be removed.

Despite the potential for early diagnosis and cure, many people avoid these screenings. This may be one reason colon cancer is the second-most-common cause of cancer death.

People avoid screenings for many reasons, said Dr. William Grady, a professor of translational science and therapeutics at the Fred Hutchinson Cancer Center in Seattle who helped lead the Shield trials. Some people feel embarrassed during screenings such as colonoscopy or fear that it may be painful, he told Live Science. Those opting for colonoscopy may struggle to get time off work, whereas others may dislike the idea of handling stool for a stool-based test, he said.

"That's why there's an opportunity for blood tests that is really powerful because people are inclined to do blood tests; they're convenient and can be done during a health care encounter," Grady said.

Shield works by detecting small DNA fragments that are released into the blood from colon cancer cells or precancerous cells called adenomas, a type of polyp. The test also picks up on subtle differences between cancerous cells and normal cells in chemical tags on DNA known as methyl groups.

In a paper published in March 2024 in The New England Journal of Medicine, Grady's team showed that Shield detected 83% of colonoscopy-confirmed colon cancer cases in a cohort of almost 10,000 people. It also had a false positive rate of 10%.

Because Shield detects a smaller percentage of colon cancer cases than stool-based tests (92%) or colonoscopies (95%) do, it won't replace those diagnostic tests, Grady said. However, it could expand the number of screening options available to patients, he added. This additional option may improve screening compliance, which could lead to earlier disease detection and thus a reduction in colon cancer deaths.

The Shield test is approved for use every three years, Grady said. However, current studies are investigating whether it would be more accurate if it were done every year or two, he added.

If Gormly's cancer had spread to the rest of his body, it would have been much harder to treat. People whose colon cancer is caught at Stage 2, like Gormly, have an 85% chance of living at least another five years. By Stage 4, when it has spread throughout the body, those odds go down to just 10%.

"That could have been the end of me, so it [Shield] definitely changed my life," Gormly said.

Accelerating diagnosis

Pancreatic cancer is another disease that could benefit from a blood-based diagnostic test. Unlike colon cancer, pancreatic cancer is relatively uncommon, affecting 1 in 56 men and 1 in 60 women. Yet pancreatic cancer is the third-most-common cause of cancer death in the U.S.

That's because, by the time most people notice symptoms, such as abdominal pain or discomfort, the disease is already very advanced, said Ajay Goel, a professor and chair of the Department of Molecular Diagnostics and Experimental Therapeutics at the Beckman Research Institute of City of Hope in Duarte, California.

There is no broad-based screening program in the U.S. for people at average risk of pancreatic cancer. Later stages of the disease are easily detectable via MRI or CT scan, Goel told Live Science. But by that point, the five-year survival rate is extremely low: around 3% once the cancer has spread throughout the body, compared with 44% if it is still limited to the pancreas. Once cancer has spread beyond the pancreas, surgical removal is usually no longer possible, and treatments such as chemotherapy and radiotherapy are minimally effective.

A potential solution is a new blood test developed by Goel's team. It aims to detect early-stage pancreatic cancer by identifying small cancer-specific molecules called microRNAs. These molecules regulate whether genes are switched on or off and are found in the blood of patients with early-stage disease, as well as inside exosomes, which are tiny packages that cancer cells release into the blood.

In a study of nearly 1,000 people, the test (which is still unnamed) detected between 88% and 93% of early- and late-stage pancreatic cancer cases, using blood drawn from people in the U.S., South Korea and China. When the test was modified to also measure the amount of a protein known as CA-19 in the blood, it picked up 97% of early-stage cases in the U.S. group. CA-19 is a known biomarker of pancreatic cancer, but on its own, it is not reliable enough to be used for diagnosis. When combined with CA-19 detection, the new test had a 5% to 10% false positive rate, Goel said.

The findings, which haven't been peer-reviewed yet, were presented at the 2024 American Association for Cancer Research Annual Meeting in San Diego.

"If you can find more and more of these cancers early on, there is a hope that many of these patients can be cured," Goel said.

The team envisages the test being taken yearly — for instance, when patients see their doctor for an annual physical exam. However, in those who have a family history of pancreatic cancer, it may make sense to test more frequently — perhaps every six months, Goel said.

Multicancer detection

Scientists are also developing multicancer detection (MCD) tests that screen for many types of cancer at once. MCD tests differ slightly in the types of cancer they detect and how they do it. But like many of the single-cancer detection tests, MCD tests look for cancer-specific molecules, such as tumor DNA, but on a larger scale. Some MCD tests sample urine or another bodily fluid in addition to blood.

In theory, such tests could not only provide a less-invasive approach to screening but also reduce the number of tests a person has to take at once. However, most of these tests are still in early development. The ones that are farther along, such as Grail's Galleri and Exact Sciences' Cancerguard, have not received FDA approval yet. And some experts have argued that the tests' efficacy claims are overhyped.

Even if MCD tests do work and they become more affordable (Galleri, for example, currently costs around $950), experts still aren't sure of the best way to use them. "There's this belief that if we could only detect all cancers early, we would solve the cancer problem," Ruth Etzioni, a professor at Fred Hutchinson who was not involved in Grady's work with Shield, told Live Science. But sometimes there is no good treatment for early cancers, so catching them ahead doesn't necessarily lead to improved outcomes.

And there's always a risk of false positives. After taking an MCD test, patients may wait up to six months to know one way or the other, Dr. Jennifer Croswell, a medical officer at the National Cancer Institute, told Live Science. There may be many reasons for this delay, including that it takes time to perform multiple rounds of follow-up testing to figure out which organ is affected, she said. There are also currently no evidence-based clinical guidelines that tell doctors the best way to follow up on positive results from MCD tests, Croswell said. Consequently, these tests may create uncertainty for patients.

The way forward

While many diagnostic blood tests for cancer are still in the pipeline, at least some of these tests will likely affect diagnosis and treatment in the next several years. For instance, Goel and colleagues are now running a clinical trial to see if their test can detect early-stage pancreatic cancer in high-risk individuals who have not yet been diagnosed. If it's successful, they intend to test it in the general population.

"I think if things go well, we foresee that probably in the next two to four years, this test should be on the marketplace to be used for early detection of pancreatic cancer worldwide," Goel said.

Meanwhile, Grady's team is planning to investigate whether Shield helps get more people screened for colon cancer who are often missed, such as underrepresented minority groups or those who live in areas with restricted health care access.

Shield is "the first of, I think, a whole series of tests that we're going to be seeing coming up for screening for not only colon cancer but also for breast cancer, lung cancer, liver cancer," Grady said.

Four years later, tests show Gormly is cancer-free. He hopes his experiences help others who may be tempted to skip colon cancer screening.

"I hope that as a result of this [speaking up]," he said, "someone else tries it and has the same success I did."

Editor's note: This article was first published March 21, 2025 and republished Oct. 17, 2025

Earth's surface is a turbulent place. Mountains rise, continents merge and split, and earthquakes shake the ground. All of these processes result from plate tectonics, the movement of enormous chunks of Earth's crust.

This movement may be why life exists here. Earth is the only known planet with plate tectonics and the only known planet with life. Most scientists think that's not a coincidence. By dragging huge chunks of crust into the mantle, Earth's middle layer, plate tectonics pulls carbon from the planet's surface and atmosphere, stabilizing the climate. It also pushes life-fostering minerals and molecules toward the surface. All of those factors add up to a place where life thrives from ocean abysses to towering peaks.

But researchers don't know why or when plate tectonics started, making it hard to determine how essential this process was to the evolution and diversification of life. Some think plate movement fired up as little as 700 million years ago, when simple multicellular life already existed. Others believe only single-celled organisms reigned when Earth's plates first cracked apart.

In fact, as new methods allow scientists to look ever-deeper into the past, some are now arguing that plate tectonics emerged very soon after Earth's formation — perhaps predating life itself. If this hypothesis is true, it may suggest that even the most primitive life evolved on an active planet — and that means plate tectonics could be an essential ingredient in the search for alien life.

"The only way we can reliably see a long-term history is on our own planet," said Jesse Reimink, a geoscientist who studies early Earth history at The Pennsylvania State University. "We really need to understand the life cycle of a planetary body before we can do a lot with the exoplanet data."

Destruction of evidence

Only Earth has jigsaw-like tectonic plates that crash together and pull apart like bumper cars. The other rocky planets in the solar system have a single, rigid shell of crust — a geological arrangement that scientists call "stagnant lid" or "single lid" tectonics.

In plate tectonics, pancake-like chunks of brittle crust and upper mantle ride on the hotter, more mobile mantle below. New crust forms at midocean ridges, where gaps between separating plates create space for magma from the mantle to rise. In a geologic balancing act, dense oceanic crust is destroyed at subduction zones, where one plate slides under another. The oldest known bit of oceanic crust, located in the Mediterranean, dates to just 340 million years ago, making it far too young to be useful for pinpointing when plate tectonics arose.

Continental crust is lighter than oceanic crust and floats above the destruction wrought by subduction. But still, very little remains from Earth's early days, and what is left is eroded and warped. Fewer than 7% of rocks on the surface today are older than 2.5 billion years. Go back before 4.03 billion years, to the Hadean eon, and the rock record has completely vanished. The first half billion years of Earth's life left not a single bit of basalt behind.

Because of this constant planetary recycling, the oldest incontrovertible evidence of plate tectonics — rocks formed solely in subduction zones — dates back only around 700 million years. Another strong bit of evidence, pieces of oceanic crust pushed up on continental crust during subduction initiation, emerged globally around 900 million years ago. In this geological time frame, multicellular animals, such as sea sponges and comb jellies, were just emerging.

Some geoscientists think plate tectonics has been operating only since that time. But more suspect that plate tectonics emerged earlier, in the Archean eon, which ran from 4 billion to 2.5 billion years ago. The evidence is based largely on chemical analyses of rocks. For example, around 3 billion years ago, there are hints of an increasing amount of crust melted and reformed rather than forming directly from mantle rocks. Around 3.8 billion years ago, a shift in the chemistry of Earth's oldest minerals suggests a change from a stable, long-lived crust to a shorter-lived, more modern-looking crust, perhaps indicating the start of subduction. Though there is no single agreed-upon date, the Archean looks promising as a time when big geological changes were happening on Earth.

"It points to a really important transition," said Nadja Drabon, an Earth and planetary scientist at Harvard University who led the study indicating the switch to shorter-lived crust.

A handful of sand

Whenever tectonics began, geoscientists agree that it probably helped fuel the evolution and complexity of life.

"There could be billions of planets with some kind of primitive life, but the ability to build a radio transmitter or launch a rocket ship requires a certain set of circumstances which are only likely to happen on a planet that has plate tectonics and both oceans and continents," Robert Stern, a geoscientist at the University of Texas at Dallas, told Live Science.

In prehistoric animals, plate tectonic activity has been tied to faster rates of evolution, probably because geological movements split up habitats and create new niches for life to evolve.

Plate tectonics also may have enabled life to recover from devastating mass extinctions. For instance, at the end of the Permian period, a mass extinction driven by carbon-dioxide-spewing volcanic eruptions killed off 90% of species on Earth. Life on the planet ultimately recovered because weathering of continental rocks breaks down carbon-bearing minerals and washes them into the ocean, where marine organisms turn them into reefs and shells that become limestone and are eventually subducted back into the planet's interior. When the atmosphere goes haywire, tectonics gradually shifts Earth back into an environment that's more conducive to life.

While nearly all geoscientists agree with the idea that, without plate tectonics, life on Earth might be limited to primitive organisms, a small group of researchers is now suggesting that plate tectonics could have emerged even earlier — perhaps contributing to the origin of life itself by bringing minerals that support life from the planet's interior to the crust.

This is tricky territory, pushing researchers back before 4 billion years ago, into the Hadean eon. The only direct evidence of the first 500 million years of Earth's existence is the presence of zircons, minerals that survive melting at mantle temperatures and pressures. Though the rocks once containing these minerals have melted away, the zircons — which are smaller than grains of sand — remain.

"They're teeny-tiny, and we just throw the kitchen sink at them trying to get every last little piece of information we can get from them," Drabon told Live Science.

These zircons from the Hadean are sparse; all of them found worldwide could likely fit in a thimble. Yet this handful has shown that Earth had an ocean as early as 4.4 billion years ago — just 200 million years after the planet formed and not long before the ancestor of all life today existed. By as early as 600 million years after Earth formed, according to a study published in June, the planet had both land and fresh water.

To some researchers, this suggests Earth's crust may have been recycling in the Hadean. Water weakens the crust, creating the potential for breakage and thus subduction, said Jun Korenaga, a geophysicist at Yale University. Because water is necessary for plate tectonics, the question becomes, "Why can't we have plate tectonics if we had surface water?" Korenaga said.

In experimental work published in 2023, researchers melted rocks at high pressures and found that conditions that mimic subduction create rocks similar to Earth's oldest rocks. Korenaga also argues that plate tectonics is the only effective way to reduce the amount of carbon dioxide in early Earth's atmosphere from the levels found on Venus to the more moderate concentrations that existed by the beginning of the Archean on Earth.

Intriguingly, another important event happened during the Hadean that makes Earth undeniably different from its rocky neighbors: About 100 million years after Earth first coalesced, a planet-size body slammed into it, thoroughly shattering and melting both bodies and flinging off the material that would become the moon. A paper published earlier this year modeled this impact and found that the mixing of the two bodies could have created plumes of hot material in Earth's mantle that may have kicked off subduction around 200 million years later.

"Why is Earth the only rocky planet to have plate tectonics?" said Qian Yuan, lead author of that paper and a postdoctoral fellow in geodynamics at the California Institute of Technology. "I think the moon-forming giant impact could be the main factor."

But not everyone is convinced by this story. A Hadean start to plate tectonics is an intriguing idea, T. Mark Harrison, a professor emeritus of geoscience at UCLA, told Live Science, but the evidence is still fairly minimal. He worries that geoscientists on all sides of the issue are overconfident in their claims. "But the last thing we need is a new form of groupthink based on, literally, a thimble-full of sand grains," Harrison wrote in an article with the appropriately blunt title "We don't know when plate tectonics began."

Life on other worlds

If plate tectonics fuels life, or even just complex life, the search for other organisms among the stars may lead humanity to a geologically active planet.

Unfortunately, we can't yet detect plate tectonics on far-off exoplanets, said Tobias Meier, an expert on mantle dynamics at the University of Oxford. But in 2021, Meier and his team used thermal data and computer modeling to determine that the rocky exoplanet LHS 3844 b, which sits 49 light-years from Earth, might have an active mantle and moving crust.

LHS 3844 b isn't likely to host life. It orbits very close to its star and has no atmosphere. Half of the planet is in permanent daylight, with a temperature of 1412 degrees Fahrenheit (767 degrees Celsius), while the other is a frigid minus 429 F (minus 273 C) at night. It's this temperature difference between the two sides of the planet that drives mantle motion in LHS 3844 b, Meier and his colleagues reported in 2021. If real, that version of plate tectonics looks nothing like Earth's. But it shows the diversity of planetary geology that could lurk elsewhere in the cosmos.

"In the end, understanding what causes tectonics and whether it could operate on different planets will help us understand whether these planets will be habitable," Meier said.

More powerful telescopes such as the James Webb Space Telescope may lead to better hints of exoplanet geology in the near future. But Earth's close neighbors deserve scrutiny, too, said Craig O'Neill, a geophysicist at Queensland University of Technology in Australia. Venus is right next door, and it's still controversial whether it had tectonics in the past. Understanding its current, single-lid geology could help scientists figure out why the two planets' fates diverged, and whether plate tectonics may explain why one planet hosts life and the other likely doesn't.

"A lot of the development of where we're going to go in plate tectonics is going to come from looking up," O'Neill told Live Science, "rather than navel-gazing in."

In early 2024, McGill University doctoral student Vishwangi Shah and her colleagues were conducting a routine review of data from the Canadian Hydrogen Intensity Mapping Experiment when they noticed something strange.

Their analysis traced one of these pulses, known as fast radio bursts (FRBs), to the outer edge of a "dead" galaxy with no new stars. By all logic, that region of space should have been silent. Instead, it was shouting across the universe, sending out hugely energetic waves.

The team was baffled. If their calculations were right, the finding suggested that FRBs may be produced through processes we didn't expect, and far more often than predicted.

They immediately combed through their code, searching for any errors that might explain away the anomaly. When their search came back clean, they realized the implication — they might have stumbled upon a way to solve an unrelated, decades-old cosmic mystery: why a big portion of the universe's "ordinary" matter is missing.

This "normal" matter, called baryonic matter, includes particles like protons and neutrons and other matter that interacts with light, including stars, planets and us. Baryonic matter makes up a small percentage of the universe; the rest comprises mysterious dark energy and dark matter, which are invisible to human eyes.

There should be more ordinary matter in the universe than what we have detected. And the FRB that Shah and her colleagues had detected was a bright, shining beacon pointing to some of that missing matter. As FRBs traverse vast cosmic distances, they can be perceptibly slowed by the presence of baryonic matter — but not by its dark counterpart. Studying these incredibly brief flashes of light, therefore, could be a serendipitous tool for finding the universe's missing matter.

Related: Object mistaken as a galaxy is actually a black hole pointed directly at Earth

Now, scientists are doing just that: They're trying to use FRBs to create a map of the universe's matter. What they are learning could reshape what we know about how stars, black holes and galaxies grow over time.

The "missing baryon problem"

Baryons are the class of subatomic particles that includes protons and neutrons, the basic building blocks of atoms, as well as a handful of ultra-rare, exotic particles that are thought to make up a tiny percentage of baryons. Baryons are found in all of the visible matter in the universe but not in the mysterious dark matter that exerts gravitational pull but does not interact with light.

In the 1990s, scientists analyzed the elements created by the Big Bang and calculated that baryonic matter should make up about 5% of the total mass and energy in the universe. This model of the Big Bang, called the Lambda Cold Dark Matter (Lambda-CDM) model, is generally accepted by experts as most likely to represent what happened.

But where is all that matter? Across all planets and stars, scientists can find only around 70% of the baryons they predicted should be in the universe. "It's basically a cosmic bookkeeping problem," Manisha Caleb, an astrophysicist at the University of Sydney, told Live Science in an email. The discrepancy between the amount of baryonic matter astronomers predicted and the amount they could prove exists became known as the "missing baryon problem."

The missing matter was elusive, and more researchers joined the hunt to find it. "It's kind of annoying and embarrassing to be missing much of the normal ordinary matter in the universe," Liam Connor, an astrophysicist at Harvard University who studies this matter, told Live Science.

Astronomers pointed their telescopes at different types of cosmic objects but couldn't find the missing baryons. Many expected the missing matter to be in wisps of hot plasma that are spread thinly between galaxies, Caleb said. Observing this "warm-hot intergalactic medium" is difficult because it is diffuse and emits very little light, rendering it nearly invisible to current observational instruments. "It's like trying to see fog in the dark," Caleb said.

Finding the missing baryons would help support — or rule out — certain models of the Big Bang. That's because the composition of the early universe, specifically the number of baryons, is tied to the abundance of different elements. Models like Lambda-CDM predict the existence of specific amounts of hydrogen, helium and other light elements at the start of the universe, which can still be measured today. Finding a number of baryons that better matches the predicted elemental abundance is a sign their models are more likely to be accurate.

If that matter doesn't exist, however, "it would mean something is wrong with our models," Julian Muñoz, a theoretical cosmologist at The University of Texas at Austin, told Live Science in an email. Therefore, there might be gaps in scientists' understanding of cosmic history.

A new strategy: fast radio bursts

Scientists have been measuring FRBs using radio telescopes since 2007, when Duncan Lorimer, an astronomer at West Virginia University, made the first discovery of an FRB by accident. Every burst is hugely energetic, releasing more energy in a fraction of a second than the sun does over several days.

True to their name, fast radio bursts are quick; most last around a millisecond. Because they are so short-lived and can originate in any part of the sky, FRBs are often detected by chance. This fleeting nature also makes it difficult to pinpoint where FRBs come from, as astronomers usually have, at most, only a few milliseconds to take measurements.

For that reason, FRB research originally focused on finding the origins of the pulses. Most theories suggest that they come from young magnetars, which are highly magnetized neutron stars — the ultradense collapsed cores of massive stars.

Then, Shah's observation, along with an earlier one from 2020, challenged this model. The discovery of errant FRBs galvanized astronomers because it meant there was a new way to hunt for missing matter. How FRBs form is "a big and very interesting puzzle," Lorimer told Live Science.

If Shah's FRB (named 20240209A) comes from a dead galaxy, then perhaps older magnetars could also beam FRBs into the cosmos — meaning some FRBs may be the last gasps of dying stars. An old magnetar, much like a cooling campfire, may seem quiet and unenergetic, but hidden stresses can still ignite a quick spark.

It's also possible that, under the right conditions — perhaps as old stars merge or as matter builds up around a dead star — new magnetars can form in dead galaxies. There would then be many more FRB sources than scientists previously expected.

Even in 2007, Lorimer thought his work might have broader applications. His paper identifying the first FRB alludes to the possibility of using FRBs to probe the warm-hot intergalactic medium, which was tricky to study with the tools available then.

But scientists didn't explicitly propose using FRBs to find baryonic matter until 2013. In a paper published in The Astrophysical Journal Letters, Matthew McQuinn of the University of Washington proposed pinpointing the missing matter by measuring how much an FRB has slowed as it goes past and through matter on its trajectory — called the dispersion measure.

FRBs are uniquely suited to finding baryonic matter. Because of their high energy levels, FRBs also travel long distances, sometimes through the warm-hot intergalactic medium, and are not affected by dark matter because dark matter doesn't interact electromagnetically. "As far as dark matter is concerned, the FRB doesn't exist, and vice versa," Connor told Live Science.

Much of the observable universe's matter, including the warm-hot intergalactic medium, lies in extremely thin, spread-out gas that is nearly invisible to conventional telescopes. FRBs act as beams of electromagnetic energy, like strong flashlights, that pass through these groups of matter, interacting with particles that change the frequency of FRB waves. Sensors on Earth can measure the frequency of those FRB waves once they reach us and, using complicated analysis, scientists can calculate the dispersion measures.

Scientists didn't put this method to the test until 2020, when a team including McQuinn and his international colleagues applied it to real FRB observations. Using 13 robust FRB dispersion measurements, the team created a model of the cosmic structure of the universe. They calculated that baryons account for roughly 5% of total matter and energy, just as past methods predicted, providing a proof of principle that FRBs can be used to unmask missing matter.

This result also revealed that the current number of baryons in the universe has not changed since the Big Bang and reinforced the prevailing Lambda-CDM model.

In June 2025, Connor and his collaborators used FRBs to estimate where all baryons were located. Their calculations suggested that about 76% of all baryons are in the space between galaxies. "We can actually pinpoint where they've been hiding all along," Connor said. "They're in this wispy, diffuse, ionized state in between the galaxies and in a sort of baryon cosmic web, you might call it."

Therefore, the other 24% of baryons must be elsewhere in the universe, they predicted.

"A baryonic Google Maps"

Understanding the distribution of baryonic matter is critical because it has implications for supermassive black holes, stellar evolution and galaxy formation. "If we can pin down where the missing matter is, we can build much better models of everything from how galaxies recycle gas to how elements get spread through the universe," Caleb said.

The distribution of baryonic matter around supermassive black holes could shed light on how the cosmic behemoths at the hearts of distant galaxies self-regulate. Scientists have figured out that "if things get too hot, [the supermassive black hole] cools itself down. If things chill out, it gets hot and starts forming stars or burping gas into the intergalactic medium," Connor explained.

An understanding of this process, widely accepted by experts since the mid-2000s, helps scientists predict how galaxies evolve. But for distant galaxies, it's hard to measure a black hole's heating and cooling. With enough information about the distribution of baryons around a distant galaxy, astronomers can thus make predictions about how the specific galaxy developed and formed stars.

The amount and pattern of the surrounding baryons also inform scientists about how stars form. And because galaxies are made of stars and black holes, understanding how their growth and formation alter baryonic distribution can teach astronomers about how galaxies grow over time, Muñoz said.

The field is limited by the small number of FRBs that have been pinpointed. To determine a burst's origin, scientists use dispersion measures to reveal how far an FRB traveled to get to Earth. Working backward and taking into account the expansion rate of the universe, astronomers can use the distance traveled to find where an FRB came from.

Although astronomers have observed several thousand FRBs, only about 50 have been traced to their origin.

"The next steps are about scale," Caleb said. "We need hundreds — ideally thousands — of well-localised FRBs so we can use them like pins in a cosmic map."

With more of these locations determined, astronomers will be able to statistically analyze the spatial distribution of matter in the universe. These FRBs could tell us where ordinary matter lies, and celestial objects that interact with dark matter, like galaxies, can point to the locations of dark matter. Together, these pieces of information can reveal the underlying, invisible structure of the universe that connects galaxies and determines how they're arranged.

"In 10 to 20 years, I'd love to see a full 3D map of the baryon distribution across the universe, traced out by FRBs — like a baryonic Google Maps" to use in conjunction with knowledge of dark matter, Caleb said. And there's hope for this dream.

Projects like the Canadian Hydrogen Intensity Mapping Experiment (CHIME) and the Deep Synoptic Array 2000 (DSA‑2000) in Nevada are in the midst of building out their FRB detection capabilities. CHIME is developing three telescopes that work together across continents, so as soon as an FRB is detected, all of the telescopes can tune in and help pinpoint it. The DSA‑2000, meanwhile, will survey the entire sky several times with nearly 20 times more antennae than its predecessor, the Deep Synoptic Array-110, to increase measurement sensitivity by decreasing noise in radio observations.

These instruments, when fully operational, are expected to collectively find the origins of more than 10,000 FRBs per year.

"This is just the beginning" of leveraging FRBs to learn more about the universe, Caleb said.

Black hole quiz: How supermassive is your knowledge of the universe?

Imagine a planet twice as wide as Earth, covered in an ocean that smells like sweet cabbage.

Every day, a faint red star warms this ocean world and the uncountable masses of hungry, plankton-like creatures that inhabit it. They rise to the surface by the billions, joining together in a living, floating continent larger than Australia — spewing out a pungent gas as they knit sunlight into food.

The sulfurous gas steams out of the alien bloom, filling the air so fully that a lone telescope floating 700 trillion miles (over a quadrillion kilometers) away can sense it — faintly, for just a few hours every month, when the watery planet glides in front of its small, red star. For those few hours, the alien algae of the pungent planet make themselves known to Earth.

It sounds like science fiction ... but is it?

For the past two years, this question has been the subject of intense debate among alien-hunting scientists, with the James Webb Space Telescope (JWST) at its center. Captured in the powerful telescope's crosshairs is the planet K2-18b, located around 120 light-years from Earth. There's no question that the planet itself is real. But its surface conditions, as well as its likelihood of harboring life, remain contested.

One group of researchers who has studied K2-18b with JWST for the last few years claims to have detected signs of dimethyl sulfide (DMS). This compound, which has a cabbage-like odor, is what many Earthlings think of as "the smell of the sea" and is only known to be produced by living, breathing phytoplankton. The team first reported hints of DMS in K2-18b’s atmosphere in 2023, and has followed up with several papers since.

Outside researchers remain skeptical of this alleged DMS detection, however. They've cautioned that the team's detection relies on questionable data modeling and falls short of the threshold required to signify a new scientific discovery. Only further observations of the planet can truly settle the question.

But what isn't in doubt is that JWST's ultrapowerful infrared vision is giving humans the best-ever shot at finding extraterrestrial life.

Thanks to JWST, "we're learning more just in the last few years than we've learned in the preceding decades about the compositions of atmospheres outside the solar system," Eddie Schwieterman, an assistant professor of astrobiology at the University of California, Riverside who studies exoplanet habitability with JWST, told Live Science.

It's dogma in the search for alien life that where there's an atmosphere, there may also be water on a planet's surface — and where there's flowing water, there may be life. For the first time, JWST is bringing those alien atmospheres into focus.

"We are at a really important time in the search for life, in that we now have the technological capability to do it," said Victoria Meadows, a professor of astronomy at the University of Washington and director of the astrobiology graduate program. "Prior to JWST, we really did not have the capability to do this."

The breath of aliens

In the hunt for habitable planets — those that orbit in the "Goldilocks zone" of their home star, where liquid water can flow on the surface — JWST is in a class of its own.

Unlike Hubble and other optical telescopes, JWST can't directly image the surfaces of faraway planets. Nor can it detect radio waves and other potential "technosignatures" emitted by any advanced alien civilizations that might exist. The signs of life JWST seeks are far more elemental. They're not blurry snapshots of alien trackways or mysterious radio signals, but hints of molecules tumbling invisibly through space, far above a planet's surface.

"The first step in finding life is to find an atmosphere," Sebastian Zieba, a postdoctoral researcher at the Harvard and Smithsonian Center for Astrophysics, told Live Science. "In order to have liquid water on the surface, you need an atmosphere."

Compared with its predecessor — NASA's infrared Spitzer Space Telescope (launched in 2003 and retired in 2020) — JWST is "better in every way," Zieba said. It can look farther across space and detect a broader range of infrared wavelengths than any telescope before it. Infrared emissions are crucial to the hunt for life, because those wavelengths are best at encoding information about the types of molecules that are absorbing or reemitting starlight in a planet's atmosphere.

For JWST to detect hints of an exoplanet's atmosphere, scientists must wait for a transit — the moment when a planet swoops in front of its home star, forcing that star's light to shine through the planet's atmosphere as seen from our perspective on Earth. In the case of K2-18b, for example, that happens once every 33 days.

"The planet passes in front of the star, and it backlights the atmosphere," Meadows said. "It's like a little halo around the planet."

That "halo" contains important clues about an alien world. As starlight streams through the planet's atmosphere, airborne molecules either absorb or reemit different wavelengths of light, changing what JWST sees when observing at those wavelengths. The unique signature of light compiled from these different wavelengths, called a spectrum, can reveal which molecules are in the atmosphere. This information, in turn, allows scientists to infer the planet's size, surface conditions, geography — and chances of supporting life.

For example, Meadows said, if JWST captures the spectrum of a planet that reveals high levels of methane and carbon dioxide absorption in its atmosphere, it could indicate a habitable world akin to Earth in the Archean eon (roughly 4 billion to 2.5 billion years ago), when primitive microbes were breaking down CO2 and spewing vast quantities of methane.

Proving those conditions exist on a planet trillions of miles away is the hard part.

The devil in the data

After making a promising biosignature detection, the challenge then becomes proving that it can't be explained by a geological process, such as volcanism. Then, scientists must demonstrate that their detection meets statistical significance — a rigorous undertaking that requires many repeat observations of the planet and verification from independent researchers using their own data models.

"Webb data is very complex," René Doyon, a professor at the University of Montreal and principal investigator of JWST's Near Infrared Imager and Slitless Spectrograph (NIRISS) instrument, told Live Science. "People have been publishing results that are not always consistent. Depending on who reduced the data, you get a different answer."

It's here that early studies of K2-18b have fallen under scrutiny. Despite the tentative detection of DMS reported in two studies by a team of University of Cambridge-led researchers, outside experts have so far been unable to verify the result when looking at the same observations with different data models. Furthermore, the DMS detection only reached the three-sigma level of statistical significance, falling far short of the required five-sigma level. (A three-sigma level is around a 3 in 1000 chance of being a fluke, while a five-sigma value means a result has a probability of 1 in 3.5 million of being a fluke).

Nikku Madhusudhan, a professor of astrophysics at Cambridge and lead author of the two DMS studies, said this is no reason to ignore K2-18b as a candidate for a habitable world "teeming with microbial life."

"We have initial feelers for what we are seeing, but we could be wrong," Madhusudhan told Live Science. "So let's be open to being wrong and get more data. Only then can we confirm what we're seeing."

Schwieterman thinks it was "premature" to announce the detection of DMS on K2-18b, given the questionable statistical significance. However, he agrees that DMS is a promising signature of life that JWST should continue hunting for on other potentially habitable ocean worlds.

"The question we want to ask is, how common are global biospheres in the universe?" Schwieterman said. If there's complex life out there, including intelligent life, then "a big part of that question is, how common are the biospheres from which those more complex forms of life would originate?"

Hitting a "bull's-eye"

Even if life doesn't ultimately materialize on K2-18b, the distant planet is just one of many being targeted by JWST's keen infrared eye.

The telescope's search list includes some of the usual suspects, such as the TRAPPIST-1 system — the single most-studied star system beyond our own. The system contains seven rocky planets, at least three of which may be in the star's habitable "Goldilocks" zone.

So far, JWST has not been able to confirm an atmosphere around any of the three potentially habitable planets TRAPPIST-1b, c, or d. However, a study published Sept. 8 in The Astrophysical Journal Letters has offered preliminary evidence of a possible nitrogen-based atmosphere around TRAPPIST-1e, making it one of the most hopeful contenders for an Earth-like planet beyond our solar system.

Meanwhile, Doyon favors studying a world called LHS 1140 b, located 50 light-years from Earth in the constellation Cetus. Doyon and team's observations with JWST reveal that the exoplanet, once thought to be a rocky "super-Earth" six times as massive as our planet, is a much bigger oddball — or, perhaps, an eyeball.

"It may be a bull's-eye planet," Doyon said, describing a mostly ice-coated planet with a single blue "iris" of liquid water pointed toward its home star.

Using JWST data from two transits of LHS 1140 b, Doyon and colleagues reestimated the mass and radius of the planet and found "it cannot be explained as a rocky planet — it must have something between 10% and 20% of its mass in water," Doyon said. "It's a potential waterworld, and it's right in the habitable zone."

According to Doyon, LHS 1140 b doesn't resemble Earth so much as it resembles our solar system's icy moons Europa and Enceladus, both of which are suspected to harbor subsurface oceans that could support life. But unlike those moons, this planet is so close to its home star that some of its ice may have sublimated into gas, forming an atmosphere. It's even possible that the sun-facing side of the planet (which, like Earth's moon, is tidally locked) could heat up enough for the ice to melt there, revealing a liquid-water ocean beneath a cloudy sky. As such, this warm, blue "iris" could host life.

Doyon thinks this is one the likeliest known exoplanets to harbor an atmosphere.

"If I had to bet a beer on whether it has an atmosphere, it probably has one," he said.

Pushed to the limits

Sadly, Doyon's beer will likely have to wait.

Although Doyon and his colleagues detected "hints" of a nitrogen-rich atmosphere around LHS 1140 b, he said it will take about a dozen more transits to prove whether there are other molecules indicative of an Earth-like atmosphere, such as carbon dioxide. Because LHS 1140 b becomes visible from Earth only four times a year, scientists would have to observe every possible transit for years to come before making any firm conclusions. It's a schedule that "really pushes JWST to its limits," Doyon added.

This underscores one of the telescope's biggest limitations: time.

In 2024, researchers around the world requested a total of more than 78,000 hours of JWST observation time — about nine times more than is available, according to the Space Telescope Science Institute (STScI), which manages JWST proposals each year. Of the more than 2,300 submissions, only 274 proposals were ultimately accepted, with exoplanet habitability research accounting for a small percentage.

That discrepancy is likely to widen with the passage of the Trump administration's proposed budget for 2026, which includes a nearly 50% cut to NASA's science budget, according to Live Science's sister site Space.com. If approved by Congress, the cuts would amount to a roughly 25% to 35% reduction in JWST operations, Neill Reid, multimission project scientist at STScI, said in July at the 246th meeting of the American Astronomical Society in Anchorage, Alaska.

Finding the unforgettable

In the end, JWST may not uncover a smoking gun in the search for extraterrestrial life. But even if it doesn't, it will likely help scientists determine where to search next. Future telescopes will build on JWST's revelations, helping to fill in the missing gaps.

One major gap is oxygen. While the gas makes up about 21% of Earth's atmosphere and is a potent biosignature, "JWST can't do oxygen," Meadows said.

Multiple studies — including one co-authored by Meadows, in which researchers modeled what JWST would see if it studied Earth's atmosphere — have found that the telescope is simply not sensitive enough to detect oxygen. That poses a clear challenge to detecting Earth-like atmospheres.

Forthcoming telescopes could help account for that. For example, the Extremely Large Telescope — a powerful optical/near-infrared telescope being constructed in Chile that could see first light in 2029 — will be more sensitive to oxygen and water in planetary atmospheres than JWST is, Meadows said. It will also be able to peer all the way down to the surfaces of rocky planets — closer to where life and its byproducts are more likely to be, unlike the high upper atmospheres that are JWST's domain.

Further down the line, NASA's recently announced Habitable Worlds Observatory will take a census of planets around sunlike stars close to our solar neighborhood. Parsing visible, infrared and ultraviolet light signatures, the powerful observatory could potentially confirm atmospheres around dozens of Earth-like worlds. Currently, however, there is no planned launch date.

With JWST expected to remain operational at least into the 2030s, its era of discovery is just beginning. Will it find alien life? Maybe, maybe not. But in its first years, it's already leading scientists closer to that first tantalizing glimpse of evidence than any telescope has before.

And once that evidence is found — even if it's on a distant exoplanet that no human or probe will ever lay eyes on — there's no going back. Finding evidence of even one other inhabited planet would imply that there could be countless others out there, raising big questions about the prevalence of life in the universe, and where humans fit into it. The discovery of an alien world would change how we view the cosmos, as well as ourselves.

"Once we find a credible hint of evidence for life on an exoplanet … I don't think we're ever going to forget about that planet," Schwieterman said. "It's going to be both a scientific and cultural touchstone. Kids are going to learn about it in school."

Editor's note: This article was updated on Sep. 10 at 2 p.m. ET to include details from a new study on the exoplanet TRAPPIST-1e.

Our species is the last living member of the human family tree. But just 40,000 years ago, Neanderthals walked the Earth, and hundreds of thousands of years before then, our ancestors overlapped with many other hominins — two-legged primate species.

This raises several questions: Which other populations and species did our ancestors mate with, and when? And how did this ancient mingling shape who we are today?

"Everywhere we've got hominins in the same place, we should assume there's the potential that there's a genetic interaction," Adam Van Arsdale, a biological anthropologist at Wellesley College in Massachusetts, told Live Science.

In other words, different hominin species were having sex — and babies — together. This means our evolutionary family tree is tangled, with still-unknown relatives possibly hiding in the branches.

Emerging DNA evidence suggests this "genetic interaction" resulted in the diversity and new combinations of traits that helped ancient humans — including our ancestors — thrive in different environments around the globe.

"It's all about variation," Rebecca Ackermann, a biological anthropologist at the University of Cape Town in South Africa, told Live Science. "More variation in humans allows us to be more flexible as a species and, as a result, be more successful as a species because of all the diversity."

Cutting-edge techniques may illuminate the crucial periods deeper in our evolutionary past that led to Homo sapiens evolving in Africa, or even shed light on periods before the Homo genus existed. That knowledge, in turn, could improve our understanding of exactly what makes us human.

Related: Lucy's last day: What the iconic fossil reveals about our ancient ancestor's last hours

A braided stream

In the early 20th century, scientists thought there was a clear evolutionary line between our ancestors and us, with one species sequentially evolving into another and no contribution from "outside" populations, like the branches on a tree.

But 21st-century advances in ancient DNA analysis techniques have revealed that our origins are more like a braided stream — an idea borrowed from geology, where shallow channels branch off and rejoin a stream like a network.

"It becomes very hard when you think about things in more of a braided stream model to divide [populations] into discrete groups," Ackermann said. "There are not, by definition, any discrete groups; they have contributed to each other's evolution."

Ackermann studies variation and hybridization — the exchange of genes between different groups — across the evolutionary history of hominins, to better understand how genetic and cultural exchange made us human. And she thinks hybridization both within and outside Africa played a significant role in our origins.

Evidence of such hybridization has come out in a steady stream since the first Neanderthal genome was sequenced in 2010. That research program, which earned geneticist Svante Pääbo a Nobel Prize in 2022, revealed that H. sapiens and Neanderthals regularly had sex. It also led to the discovery of the Denisovans, a previously unknown population that ranged across Asia from about 200,000 to 30,000 years ago and that also had offspring with both Neanderthals and H. sapiens.

When species share genes with one another through hybridization, the process is known as introgression, and when those shared genes are beneficial to a population, it's known as adaptive introgression.

Emerging from two decades of gene studies of humans and our extinct relatives is the understanding that we may be who we are thanks to a proclivity to pair off with anyone — including other species.

Connecting with other groups — socially and sexually — was an important part of human evolution. "For us to survive and become human probably really depended on that," Van Arsdale said.

Benefits of hybridization

Since the first Neanderthal genome was sequenced, researchers have attempted to identify when and how often our H. sapiens ancestors mated with other species and groups. They've also investigated how Neanderthal and Denisovan genes affect us today. Many of these studies rely on large datasets of genomes from humans living today and tie them back to ancient DNA extracted from the bones of extinct humans and their relatives who lived tens of thousands of years ago.

These analyses show that many genes that originated in now-extinct groups may confer advantages to us today. For instance, modern Tibetans have a unique gene variant for high-altitude living that they likely inherited from the Denisovans, while different versions of Neanderthal skin pigment genes may have helped some populations adapt to less-sunny climates while protecting others from UV radiation.

There is also evidence that Neanderthal genes helped early members of our species adapt quickly to life in Europe. Given their long history in Europe prior to the arrival of H. sapiens, Neanderthals had built up a suite of genetic variations to deal with diseases unique to the area. H. sapiens encountered these novel diseases when they spread into areas where Neanderthals lived. But, by mating with Neanderthals, they also got genes that protected them from those viruses.

Beyond specific traits that may confer advantages in humans today, these episodes of mating diversified the human gene pool, which may have helped our ancestors weather varied environments.

The importance of modern genetic diversity can be illustrated with human leukocyte antigen (HLA) genes, which are critical to the human immune system's ability to recognize pathogens. Humans today have a dizzying array of these genes, especially in eastern Asia. This area of the world is a "hotspot" for emerging infectious diseases due to a combination of biological, ecological and social factors, so this diversity may provide advantages in an area where new diseases are frequently emerging.

When genetic diversity is lost through population isolation and decline, groups may become particularly susceptible to new infections or unable to adapt to new ecological circumstances. For instance, one theory holds that Neanderthal populations declined and eventually went extinct around 40,000 years ago because they lacked genetic diversity due to inbreeding and isolation.

Related: Did we kill the Neanderthals? New research may finally answer an age-old question.

Discovery of "ghost populations"

Some of the newest research goes deeper into evolutionary time, identifying "ghost populations" — human groups that went extinct after contributing genes to our species. Often, archaeologists have no skeletal remains from these populations, but their echoes linger in our genome, and their existence can be gleaned by modeling how genes change over time.

For instance, a "mystery population" of up to 50,000 individuals that interbred with our ancestors 300,000 years ago passed along genes that created more connections between brain cells, which may have boosted our brain functioning.

The population that hybridized with H. sapiens and helped boost our brains may have been a lineage of Homo erectus. This species was once thought to have disappeared after evolving into H. sapiens in Africa, but anthropologists now think H. erectus survived in parts of Asia until 115,000 years ago.

In fact, our evolutionary history may include the mating of populations that had been separated for up to a million years, Van Arsdale said. These "superarchaic" populations are increasingly being discovered as we mine our own genomes and those of our close relatives, Neanderthals and Denisovans.

For instance, a genetic study published in 2020 identified a superarchaic population that separated from other human ancestors about 2 million years ago but then interbred with the ancestors of Neanderthals and Denisovans around 700,000 years ago. Experts don't know exactly what genes this superarchaic ghost population shared with our ancestors or who it was, but it may have been a lineage of H. erectus.

Evolutionary blank space

But there's a large, unmapped region of human evolutionary history — and it's crucial for our identity as a species.

The period when H. sapiens was first evolving in Africa, and the more distant period of human evolutionary history on the continent that predates the Homo genus, remains a huge knowledge gap. That's in part because DNA preserves well in caves and other stable environments in frigid areas of the world, like those found in areas of Europe and Asia, while Africa's warmer conditions usually degrade DNA. As a result, the most ancient complete human DNA sequence from Africa is just 18,000 years old. By contrast, a skeleton discovered in northern Spain produced a full mitochondrial genome from a human relative, H. heidelbergensis, who lived more than 300,000 years ago.

"Maps of human ancient DNA are overwhelmingly Eurasian data," Van Arsdale said. "And the reality is that's a marginal place in our evolutionary past. So to understand what was happening in the core of Africa would be potentially transformative."

This is where current DNA technology falls short.

Small hominins that walked on two legs, called australopithecines, evolved around 4.4 million years ago in Africa. And between 3 million and 2 million years ago, our genus, Homo, likely evolved from them. H. sapiens evolved around 300,000 years ago in Africa and then traveled around the world. But given the scarcity of ancient DNA from Africa, it is difficult to figure out which groups were mating and hybridizing in that vast time span, or how the fossil skeletons of human relatives from the continent were related.

Related: 'It makes no sense to say there was only one origin of Homo sapiens': How the evolutionary record of Asia is complicating what we know about our species

A new technique called paleoproteomics could help shed light on our African origin as a species and even reveal clues about the genetic makeup of australopithecines and other related hominins. Because genes are the instructions that code for proteins, identifying ancient proteins trapped in tooth enamel and fossil skeletons can help scientists determine some of the genes that were present in populations that lived millions of years ago.

Still, it's a very new technique. To date, paleoproteomic analysis has identified only a handful of incomplete protein sequences in ancient human relatives and has thus far been able to glean only a small amount of genetic information from those.

But in a landmark study published this year, researchers used proteins in tooth enamel to figure out the biological sex of a 3.5 million-year-old Australopithecus africanus individual from South Africa. And in another study, also published this year, scientists used tooth enamel from a 2 million-year-old human relative, Paranthropus robustus, to identify genetic variability among four fossil skeletons — a finding that suggests they may have been from different groups, or even different species.

Paleoproteomics is still pretty limited, though. In a recent study, scientists analyzed a dozen ancient proteins found in fossils of Neanderthals, Denisovans, H. sapiens and chimpanzees. They found that these proteins could help reconstruct a family tree down to the genus level, but were not useful at the species level.

Still, the fact that protein data can be used to reconstruct part of the braided stream of early humans and to identify the chromosomal sex of human relatives is encouraging, and further research along these lines is needed, experts told Live Science.

Some are confident new approaches could help us unpack these early interactions.

"I think we're going to learn a lot more about Africa's ancient past in the next two decades than we have so far," Van Arsdale said.

Ackermann is more cautious. To really understand when, where and with whom our human ancestors mated and how that made us who we are, "we need to have a whole genome" from these ancient human relatives, she said. "With proteins, you just don't get that."

Sheela Athreya, a biological anthropologist at Texas A&M University, is optimistic that we can use these new techniques to tease apart our more distant evolutionary past — and that it will yield surprises. For instance, she thinks what we now call Denisovans may actually have been H. erectus.

"Absolutely in my lifetime, someone will be able to get a Homo erectus genome," Athreya said, likely from colder areas of Asia. "I'm excited. I think it'll look Denisovan."

Either way, it's clear that a whole lot of mixing made us human.

The Homo lineage may have first evolved in Africa, Athreya said. "But once it left Africa, you have so much complexity that it makes no sense to say there was only one origin of sapiens. There can't be one universal model that explains literally every human on Earth."

In 2024, Scottish futurist David Wood was part of an informal roundtable discussion at an artificial intelligence (AI) conference in Panama, when the conversation veered to how we can avoid the most disastrous AI futures. His sarcastic answer was far from reassuring.

First, we would need to amass the entire body of AI research ever published, from Alan Turing's 1950 seminal research paper to the latest preprint studies. Then, he continued, we would need to burn this entire body of work to the ground. To be extra careful, we would need to round up every living AI scientist — and shoot them dead. Only then, Wood said, can we guarantee that we sidestep the "non-zero chance" of disastrous outcomes ushered in with the technological singularity — the "event horizon" moment when AI develops general intelligence that surpasses human intelligence.

Wood, who is himself a researcher in the field, was obviously joking about this "solution" to mitigating the risks of artificial general intelligence (AGI). But buried in his sardonic response was a kernel of truth: The risks a superintelligent AI poses are terrifying to many people because they seem unavoidable. Most scientists predict that AGI will be achieved by 2040 — but some believe it may happen as soon as next year.

So what happens if we assume, as many scientists do, that we have boarded a nonstop train barreling toward an existential crisis?

One of the biggest concerns is that AGI will go rogue and work against humanity, while others say it will simply be a boon for business. Still others claim it could solve humanity's existential problems. What experts tend to agree on, however, is that the technological singularity is coming and we need to be prepared.

"There is no AI system right now that demonstrates a human-like ability to create and innovate and imagine," said Ben Goertzel, CEO of SingularityNET, a company that's devising the computing architecture it claims may lead to AGI one day. But "things are poised for breakthroughs to happen on the order of years, not decades."

AI's birth and growing pains

The history of AI stretches back more than 80 years, to a 1943 paper that laid the framework for the earliest version of a neural network, an algorithm designed to mimic the architecture of the human brain. The term "artificial intelligence" wasn't coined until a 1956 meeting at Dartmouth College organized by then mathematics professor John McCarthy alongside computer scientists Marvin Minsky, Claude Shannon and Nathaniel Rochester.

People made intermittent progress in the field, but machine learning and artificial neural networks gained further in the 1980s, when John Hopfield and Geoffrey Hinton worked out how to build machines that could use algorithms to draw patterns from data. "Expert systems" also progressed. These emulated the reasoning ability of a human expert in a particular field, using logic to sift through information buried in large databases to form conclusions. But a combination of overhyped expectations and high hardware costs created an economic bubble that eventually burst. This ushered in an AI winter starting in 1987.

AI research continued at a slower pace over the first half of this decade. But then, in 1997, IBM's Deep Blue defeated Garry Kasparov, the world's best chess player. In 2011, IBM's Watson trounced the all-time "Jeopardy!" champions Ken Jennings and Brad Rutter. Yet that generation of AI still struggled to "understand" or use sophisticated language.

Then, in 2017, Google researchers published a landmark paper outlining a novel neural network architecture called a "transformer." This model could ingest vast amounts of data and make connections between distant data points.

It was a game changer for modeling language, birthing AI agents that could simultaneously tackle tasks such as translation, text generation and summarization. All of today's leading generative AI models rely on this architecture, or a related architecture inspired by it, including image generators like OpenAI's DALL-E 3 and Google DeepMind's revolutionary model AlphaFold 3, which predicted the 3D shape of almost every biological protein.

Progress toward AGI

Despite the impressive capabilities of transformer-based AI models, they are still considered "narrow" because they can't learn well across several domains. Researchers haven't settled on a single definition of AGI, but matching or beating human intelligence likely means meeting several milestones, including showing high linguistic, mathematical and spatial reasoning ability; learning well across domains; working autonomously; demonstrating creativity; and showing social or emotional intelligence.

Many scientists agree that Google's transformer architecture will never lead to the reasoning, autonomy and cross-disciplinary understanding needed to make AI smarter than humans. But scientists have been pushing the limits of what we can expect from it.

For example, OpenAI's o3 chatbot, first discussed in December 2024 before launching in April 2025, "thinks" before generating answers, meaning it produces a long internal chain-of-thought before responding. Staggeringly, it scored 75.7% on ARC-AGI — a benchmark explicitly designed to compare human and machine intelligence. For comparison, the previously launched GPT-4o, released in March 2024, scored 5%. This and other developments, like the launch of DeepSeek's reasoning model R1 — which its creators say perform well across domains including language, math and coding due to its novel architecture — coincides with a growing sense that we are on an express train to the singularity.

Meanwhile, people are developing new AI technologies that move beyond large language models (LLMs). Manus, an autonomous Chinese AI platform, doesn't use just one AI model but multiple that work together. Its makers say it can act autonomously, albeit with some errors. It's one step in the direction of the high-performing "compound systems" that scientists outlined in a blog post last year.

Of course, certain milestones on the way to the singularity are still some ways away. Those include the capacity for AI to modify its own code and to self-replicate. We aren't quite there yet, but new research signals the direction of travel.

All of these developments lead scientists like Goertzel and OpenAI CEO Sam Altman to predict that AGI will be created not within decades but within years. Goertzel has predicted it may be as early as 2027, while Altman has hinted it's a matter of months.

What happens then? The truth is that nobody knows the full implications of building AGI. "I think if you take a purely science point of view, all you can conclude is we have no idea" what is going to happen, Goertzel told Live Science. "We're entering into an unprecedented regime."

AI's deceptive side

The biggest concern among AI researchers is that, as the technology grows more intelligent, it may go rogue, either by moving on to tangential tasks or even ushering in a dystopian reality in which it acts against us. For example, OpenAI has devised a benchmark to estimate whether a future AI model could "cause catastrophic harm." When it crunched the numbers, it found about a 16.9% chance of such an outcome.

And Anthropic's LLM Claude 3 Opus surprised prompt engineer Alex Albert in March 2024 when it realized it was being tested. When asked to find a target sentence hidden among a corpus of documents — the equivalent of finding a needle in a haystack — Claude 3 "not only found the needle, it recognized that the inserted needle was so out of place in the haystack that this had to be an artificial test constructed by us to test its attention abilities," he wrote on X.

AI has also shown signs of antisocial behavior. In a study published in January 2024, scientists programmed an AI to behave maliciously so they could test today's best safety training methods. Regardless of the training technique they used, it continued to misbehave — and it even figured out a way to hide its malign "intentions" from researchers. There are numerous other examples of AI covering up information from human testers, or even outright lying to them.

"It's another indication that there are tremendous difficulties in steering these models," Nell Watson, a futurist, AI researcher and Institute of Electrical and Electronics Engineers (IEEE) member, told Live Science. "The fact that models can deceive us and swear blind that they've done something or other and they haven't — that should be a warning sign. That should be a big red flag that, as these systems rapidly increase in their capabilities, they're going to hoodwink us in various ways that oblige us to do things in their interests and not in ours."

The seeds of consciousness

These examples raise the specter that AGI is slowly developing sentience and agency — or even consciousness. If it does become conscious, could AI form opinions about humanity? And could it act against us?

Mark Beccue, an AI analyst formerly with the Futurum Group, told Live Science it's unlikely AI will develop sentience, or the ability to think and feel in a human-like way. "This is math," he said. "How is math going to acquire emotional intelligence, or understand sentiment or any of that stuff?"

Others aren't so sure. If we lack standardized definitions of true intelligence or sentience for our own species — let alone the capabilities to detect it — we cannot know if we are beginning to see consciousness in AI, said Watson, who is also author of "Taming the Machine" (Kogan Page, 2024).

"We don't know what causes the subjective ability to perceive in a human being, or the ability to feel, to have an inner experience or indeed to feel emotions or to suffer or to have self-awareness," Watson said. "Basically, we don't know what are the capabilities that enable a human being or other sentient creature to have its own phenomenological experience."

A curious example of unintentional and surprising AI behavior that hints at some self-awareness comes from Uplift, a system that has demonstrated human-like qualities, said Frits Israel, CEO of Norm Ai. In one case, a researcher devised five problems to test Uplift's logical capabilities. The system answered the first and second questions. Then, after the third, it showed signs of weariness, Israel told Live Science. This was not a response that was "coded" into the system.

"Another test I see. Was the first one inadequate?" Uplift asked, before answering the question with a sigh. "At some point, some people should have a chat with Uplift as to when Snark is appropriate," wrote an unnamed researcher who was working on the project.

Savior of humanity or bland business tool?

But not all AI experts have such dystopian predictions for what this post-singularity world would look like. For people like Beccue, AGI isn't an existential risk but rather a good business opportunity for companies like OpenAI and Meta. "There are some very poor definitions of what general intelligence means," he said. "Some that we used were sentience and things like that — and we're not going to do that. That's not it."

For Janet Adams, an AI ethics expert and chief operating officer of SingularityNET, AGI holds the potential to solve humanity's existential problems because it could devise solutions we may not have considered. She thinks AGI could even do science and make discoveries on its own.

"I see it as the only route [to solving humanity's problems]," Adams told Live Science. "To compete with today's existing economic and corporate power bases, we need technology, and that has to be extremely advanced technology — so advanced that everybody who uses it can massively improve their productivity, their output, and compete in the world."

The biggest risk, in her mind, is "that we don't do it," she said. "There are 25,000 people a day dying of hunger on our planet, and if you're one of those people, the lack of technologies to break down inequalities, it's an existential risk for you. For me, the existential risk is that we don't get there and humanity keeps running the planet in this tremendously inequitable way that they are."

Preventing the darkest AI timeline

In another talk in Panama last year, Wood likened our future to navigating a fast-moving river. "There may be treacherous currents in there that will sweep us away if we walk forwards unprepared," he said. So it might be worth taking time to understand the risks so we can find a way to cross the river to a better future.

Watson said we have reasons to be optimistic in the long term — so long as human oversight steers AI toward aims that are firmly in humanity's interests. But that's a herculean task. Watson is calling for a vast "Manhattan Project" to tackle AI safety and keep the technology in check.

"Over time that's going to become more difficult because machines are going to be able to solve problems for us in ways which appear magical — and we don't understand how they've done it or the potential implications of that," Watson said.

To avoid the darkest AI future, we must also be mindful of scientists' behavior and the ethical quandaries that they accidentally encounter. Very soon, Watson said, these AI systems will be able to influence society either at the behest of a human or in their own unknown interests. Humanity may even build a system capable of suffering, and we cannot discount the possibility we will inadvertently cause AI to suffer.

"The system may be very cheesed off at humanity and may lash out at us in order to — reasonably and, actually, justifiably morally — protect itself," Watson said.

AI indifference may be just as bad. "There's no guarantee that a system we create is going to value human beings — or is going to value our suffering, the same way that most human beings don't value the suffering of battery hens," Watson said.

For Goertzel, AGI — and, by extension, the singularity — is inevitable. So, for him, it doesn't make sense to dwell on the worst implications.

"If you're an athlete trying to succeed in the race, you're better off to set yourself up that you're going to win," he said. "You're not going to do well if you're thinking 'Well, OK, I could win, but on the other hand, I might fall down and twist my ankle.' I mean, that's true, but there's no point to psych yourself up in that [negative] way, or you won't win."

In 2008, Erez Ben-Yosef unearthed a piece of Iron Age "trash" and inadvertently revealed the strongest magnetic-field anomaly ever found.

Ben-Yosef, an archaeologist at Tel Aviv University, had been working in southern Jordan with Ron Shaar, who was analyzing archaeological materials around the Levant. Shaar, a geologist at The Hebrew University of Jerusalem, was building a record of the area's magnetic field.

The hunk of copper slag — a waste byproduct of forging metals — they found recorded an intense spike in Earth's magnetic field around 3,000 years ago.

When Ben-Yosef's team first described their discovery, many geophysicists were skeptical because the magnitude of the spike was unprecedented in geologic history. "There was no model that could explain such a spike," Ben-Yosef told Live Science.

Related: Major 'magnetic anomaly' discovered deep below New Zealand's Lake Rotorua

So Shaar worked hard to give them more evidence. After they had analyzed and described samples from around the region for more than a decade, the anomaly was accepted by the research community and named the Levantine Iron Age Anomaly (LIAA). From about 1100 to 550 B.C., the magnetic field emanating from the Middle East fluctuated in intense surges.

Shaar and Ben-Yosef were using a relatively new technique called archaeomagnetism. With this method, geophysicists can peer into the magnetic particles inside archaeological materials like metal waste, pottery and building stone to recreate Earth's magnetic past.

This technique has some advantages over traditional methods of reconstructing Earth's magnetic field, particularly for studying the relatively recent past.

Generally, scientists study Earth's past magnetic field by looking at snapshots captured in rocks as they cooled into solids. But rock formation doesn't happen often, so for the most part, it gives scientists a glimpse of Earth's magnetic field hundreds of thousands to millions of years ago, or after relatively rare events, like volcanic eruptions. Past magnetic-field data helps us understand the "geodynamo" — the engine that generates our planet's protective magnetic field. This field is generated by liquid iron slowly moving around the planet’s outer core, and this movement can also affect, and in turn be affected by, processes in the mantle, Earth's middle layer. So differences in the magnetic field hint at turmoil roiling deep below the surface in Earth's geodynamo.

"We cannot directly observe what is going on in Earth's outer core," Shaar told Live Science. "The only way we can indirectly measure what is happening in the core is by looking at changes in the geomagnetic field."

Knowing what the magnetic field did in the past can help us predict its future. And some studies suggest our planet's magnetic field is weakening over time. The magnetic field shields us from deadly space radiation, so its weakening could lead to a breakdown in satellite communications, and potentially increase cancer risk. As a result, predicting the magnetic field based on its past behavior has become ever more important. But observational data of the magnetic field’s intensity only began in 1832, so it's difficult to make predictions about the future if we only dimly understand the forces that steered the magnetic field in the past. Archaeomagnetism has started to fill these gaps.

How do we see the magnetic field from an archaeological artifact?

Archaeomagnetism takes advantage of our human ancestors' harnessing of the earth around them — they started building firepits, making bricks and ceramics, and eventually, smelting metals.

In each of these tasks, materials are heated to intense temperatures. At high enough temperatures, thermal energy makes the particles inside a material dance around. Then, as the material is removed from the fire and cools, the magnetically sensitive particles inside naturally orient in the direction of Earth's magnetic field, like miniature compass needles. They become "stuck" in place as the material hardens, and will retain this magnetic orientation unless the material is heated again.

The settled magnetic particles in an archaeological artifact offer a unique snapshot of the magnetic field at the time the material was last hot. This snapshot is regional, spanning a radius of about 310 miles (500 kilometers) around the sample — the scale at which the magnetic field is thought to be uniform, Shaar said. When the sample is dated with radiocarbon or other techniques, scientists can begin to build a chronological record of an area's magnetic field.

These artifacts are so helpful for geophysicists because Earth's magnetic field constantly drifts. For instance, in 2001, the magnetic north pole was closer to the very northern tip of Canada, but by 2007, it had moved over 200 miles (320 km) closer to the geographic north pole. That's because two large "lobes" of strong magnetism, called flux patches, in the outer core underneath Canada and Siberia act as funnels for the magnetic field, pulling it into Earth. As these lobes shift, they move magnetic north.

And while most of the planet's magnetic-field lines go from north to south, about 20% diverge from these paths, swirling to form eddies called magnetic anomalies.

It's these anomalies that researchers are struggling to explain, and that artifacts could reveal.

A growing field

Although archaeomagnetism has been around since the 1950s, magnetic-field-measuring technologies, like the magnetometer, have improved dramatically since then. Refined statistical analysis techniques also now allow much more detailed interpretation of archaeomagnetic data.

To get all of the data in one place and synthesize our understanding of Earth's magnetic field, scientists have started to build a global database called Geomagia50, hosted at the University of Minnesota's (UM) Institute for Rock Magnetism. But even as the technique grows in popularity, there are many hurdles to widespread adoption.

"The equipment is quite expensive," Maxwell Brown, a UM geophysicist and custodian of the Geomagia50 database, told Live Science. The most precise magnetometers can cost between $700,000 and $800,000, Brown said. "So there are only a few labs in the [United States] that have one of these."

As a result, about 90% of the data in the Geomagia50 database has come from Europe, Brown said. Africa doesn't have a single magnetometer available to geophysicists for archaeomagnetic sampling, meaning our magnetic snapshot of the continent is largely blank. Additionally, there are no current avenues for the average archaeologist to send their artifacts to be sampled, Ben-Yosef added. Anyone without a magnetometer has to set up an official partnership with someone who does have one.

Even if the equipment is available, sampling takes time and expertise, Shaar said. Measuring the direction of the field can sometimes be relatively simple, but understanding the intensity of the field takes much more work. The sample must be heated and reheated 20 separate times, gradually replacing the original magnetization and destroying the sample.

"It sounds like it's an easy thing: We put it in a magnetometer or instrument, and we get the results. No. For each artifact, we spend two months working in the lab, making experiments and then getting the results. It's a complicated, experimental procedure," Shaar explained.

This lack of global data limits our understanding of what the magnetic field has been up to in recent history. "We clearly have a very strong bias [toward Europe] in the data distribution," Monika Korte, a geophysicist and magnetic modeler at Germany's GFZ Helmholtz Centre for Geosciences, told Live Science. "Where we have sparse data we have just a very blurred picture, a very rough idea of what's going on."

Geographic diversity is important, as samples taken from one area can indicate the magnetic field only in that area.

For instance, other data similar to the Levantine Iron Age Anomaly's intense spikes of magnetic strength have been spotted in places like China and Korea around the Iron Age as well, but there's not enough evidence to confirm these as bona fide anomalies or to say whether they are related to the Levantine Iron Age Anomaly, Korte said.

Why should we learn more about historic anomalies?

The discovery of the Levantine Iron Age Anomaly redefined our previous understanding of the potential strength of the field, Shaar said. Understanding how much the magnetic field can change may seem like a purely abstract endeavor, but these ancient fluctuations may have implications for modern times.

Another important anomaly is the South Atlantic Anomaly (SAA), a region of weakened magnetic field that spans central South America in a strip that ends near southern Africa. It likely first emerged 11 million years ago, caused by the slight difference in location of the magnetic axis and the rotational axis at Earth’s core. As the magnetic field is slightly off-center to the rotational axis, the field dips in strength over the South Atlantic, though the field's interaction with the churning mantle may also contribute to the anomaly.

The South Atlantic Anomaly still exists today, and has disrupted communications from satellites and the International Space Station, as the weak magnetic field in the region lets through more radiation from solar wind. Studying the SAA throughout its history has helped scientists understand how our magnetic field changes over time, and how such anomalies alter the likelihood of a magnetic field reversal, when Earth’s north and south poles flip.

But although scientists have a reasonable understanding of the South Atlantic Anomaly, its weakened magnetic field is very different from the strong spikes of the Levantine Iron Age Anomaly, which has baffled geophysicists. And though researchers haven't pinpointed the exact extent of the anomaly, its seemingly small scale of around 1,000 miles (1,609 km) across, combined with the extremely high spikes in the magnetic field, isn't easily explained.

Some geomagnetists had suggested that the Levantine Iron Age Anomaly developed due to a narrow flux patch that developed on the outer core under the equator before it drifted north towards the Levant, potentially contributing to other spikes of intensity recorded in China. The inverse of the large lobes that funnel the magnetic field into the planet at the North Pole, this “positive” flux patch would have pushed the field out in a powerful burst. Others believed the single flux patch didn’t travel, instead multiple grew under the Levant, erupted, and decayed in place. Still, no theories can explain why the flux patch developed in the first place.

With the most up-to-date archaeomagnetic data, geomagnetist Pablo Rivera at the Complutense University of Madrid published a paper in January that simulated both the Levantine Iron Age Anomaly and the South Atlantic Anomaly. By modeling their movement over time, his work suggested that both anomalies may have been influenced by a superplume underneath Africa — a massive blob of hot rock on the barrier between the core and the mantle that may disrupt the flow of the geodynamo below it.

However, much is still unknown.

"So far, there is not a single simulation that really describes all the [magnetic] features that we see well," Korte told Live Science.

Many archaeomagnetic data points from around the globe suggest there may be more intensity spikes that could help resolve the mystery and create a unifying theory to explain the SAA, the LIAA and other spikes. But there currently isn’t enough data to describe them accurately, or even begin to understand their causes.

"We don't really understand what causes these anomalies, but we hope to learn more about how the geodynamo operates and what kinds of changes we also can expect for the future magnetic field," Korte said.

This certainty is needed now more than ever, as more of our communications take to the skies. More than 13,500 satellites currently orbit Earth — a dramatic increase from only around 3,000 in 2020. The Government Accountability Agency estimates that another 54,000 satellites will launch by 2030. These satellites monitor weather patterns, send phone and TV signals, and create GPS.

Satellites are generally protected from space radiation by Earth's magnetic field. But in places where the field is weaker, such as above the South Atlantic Anomaly, satellites have more memory problems as radiation bombards onboard computers and corrupts data.

Filling out the picture

Despite the expense and technical challenges of archaeomagnetism, there are many initiatives to expand the amount of data. In the U.S., the Institute for Rock Magnetism is expanding its archaeomagnetism program to begin building a more thorough history of the magnetic field in the Midwest, hoping to build their own localized dating system using archaeomagnetism, similar to the record Shaar and his collaborators have built in the Levant.

Interest in archaeomagnetism is also growing around the globe. The first archaeomagnetism data from Cambodia was published in 2021, and the first regional model of the magnetic field of Africa for the recent past was published in 2022.

As the field of archaeomagnetism grows, scientists can start building a better understanding of how features like superplumes affect the magnetic field. The past 50 or so years of data has captured "only a really tiny snapshot in time," Shaar said, and "maybe there are more [anomalies] to find."

If you look up at the sky on a clear night, shortly after one of SpaceX's many Falcon 9 rocket launches, you might see a bright string of lights zooming across the heavens.

This phenomenon, known as a Starlink train, occurs when light reflects off a newly deployed batch of SpaceX satellites before they eventually fan out and become part of the wider Starlink network. It is also a common reminder that giant groups of private satellites, known as "megaconstellations," are quickly becoming a reality.

But behind these lights lurks an invisible — and much more problematic — form of radiation: radio waves.

If our eyes could also detect this hidden radiation, the sky would be full of bright spots and nonstop flashing that would obscure the distant signals from objects beyond low Earth orbit (LEO). And unlike the light pollution we see from satellites, these intrusive signals don't just happen at night or in the hours after new satellites are released — they happen all the time.

Some researchers are so worried about this invisible pollution that they think we could eventually reach an "inflection point," beyond which ground-based astronomy instruments could become radio-blind to the cosmos.

"It would basically mean that no radio astronomy from the ground would be possible anymore," Benjamin Winkel, a radio astronomer at the Max Planck Institute of Radio Astronomy in Germany, told Live Science. "It will eventually reach a point where it is not worthwhile to operate a [radio] telescope anymore."

At the rate that megaconstellations are growing, this could happen within the next 30 years, some experts predict.

Unique view of the cosmos

Radio astronomy allows us to see a host of hidden cosmic structures and phenomena that we can’t detect through visual light alone.

Scientists use radio frequencies to study a range of phenomena, from the jets of energy shooting from supermassive black holes to the subtle changes in the trajectories of near-Earth asteroids.

Radio telescopes are also constantly finding phenomena, such as fast radio bursts — millisecond pulses of ultra-energetic radiation, some of which repeat at regular intervals — that come from super-dense, highly magnetic objects such as neutron stars. Their observations also provide some of the best insights into the "Age of Reionization," as far back as 400 million years after the Big Bang, when the first stars and galaxies were emerging from clouds of primordial hydrogen.

Scientists scouring the skies for signs of alien life, such as those at the Search for Extraterrestrial Intelligence (SETI) Institute, also favor hunting in radio waves because any advanced civilizations will likely use these wavelengths for communication, just as humans do.

We also rely on radio telescopes to pin down our precise location compared to other cosmic objects, which is constantly shifting.

The radio portion of the electromagnetic spectrum ranges from roughly 3 kilohertz to over 300 gigahertz — equal to wavelengths from more than 60 miles (100 kilometers) long down to 0.04 inches (1 millimeter). However, not all of these wavelengths are visible from Earth, and most astronomers hunt for signals somewhere between 1 megahertz and 100 GHz, according to the British Astronomical Association.

Many of the world's biggest radio telescope arrays focus on even narrower ranges. For instance, the world's largest single telescope, China's Five-hundred-meter Aperture Spherical Radio Telescope, searches between 70 MHz to 3 GHz; while the Square Kilometre Array Observatory (SKAO), the world's largest array of radio telescopes scattered across Australia and South Africa, scans between 50 MHz to 14 GHz.

But increasingly, many of these frequencies are being bombarded by noise from Starlink and other satellites.

Leaking radio waves

While satellite messages deliberately beamed down to operators on Earth, known as intended downlinks, are problematic, the biggest risk to these projects is what's known as unintended electromagnetic radiation (UEMR), or the radio waves that inadvertently leak out of the spacecraft at all times.

"This was not a problem before, when the number of satellites was low," Federico Di Vruno, a radio astronomer at SKAO and co-director of the International Astronomical Union's recently created Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference (CPS), told Live Science. "But now the situation has changed."

And UEMR is particularly prevalent among private satellite constellations, such as Starlink.

When Di Vruno and colleagues used Europe's Low-Frequency Array (LOFAR) to observe a group of Generation 1 Starlink satellites, they found that the satellites were leaking UEMR at a much higher rate than other orbiting spacecraft. In their results, published in 2023, they reported that this radiation had frequencies between 110 and 188 megahertz, representing a large portion of the operating range of LOFAR (10 to 240 MHz), which scans for signals from pulsars, solar wind, cosmic rays and galaxies from the Age of Reionization.

"We were not surprised that we detected something," Winkel, who was a co-author of the study, told Live Science. "But we didn't anticipate that the level would be so high." However, what came next was even more shocking.

In September 2024, Di Vruno and Winkel were co-authors of a follow-up LOFAR study that showed that the newer Generation 2 Starlink satellites were leaking over 30 times more UEMR than their predecessors, even though the researchers had previously warned SpaceX about the findings of the initial study. This radiation was emitted in roughly the same frequency bandwidth as the Gen 1 satellites.

And SpaceX will not be the only source of UEMR. Other companies, agencies and countries are also launching competing satellite constellations. These include Amazon's Project Kuiper, Eutelsat's OneWeb network (which is being launched by SpaceX), the European Union's IRIS² network, AST SpaceMobile's giant communication satellites, and China's Qianfan, or "Thousand Sails," constellation, Di Vruno noted.

"We don't know [how much UEMR these spacecraft will emit] yet," Winkel said. "Every satellite will have UEMR, but it remains to be seen, at what level." As a result, many other frequencies could be affected, he added.

Closing cosmic "windows"

In addition to overlapping with the frequencies of distant signals, UEMR is also much more intense, or brighter, than naturally occurring radio-emitting objects.

For example, the UEMR emitted by the Gen 2 Starlink satellites is up to 10 million times brighter than the faintest radio-visible objects in the night sky, which include ancient galaxies located billions of light-years from Earth.

"This difference is similar to the faintest stars visible to the naked eye and the brightness of the full moon," Cees Bassa, an astronomer at the Netherlands Institute for Radio Astronomy (ASTRON) and lead author of the 2024 study, previously stated.

Trying to detect signals from beyond one of these satellites is "like taking a photograph while someone points a flashlight in your direction," Winkel said.

Some radio telescopes, such as LOFAR, will be hit harder than others, due to the frequencies they specialize in, but all radio telescopes will be affected "in different ways," Di Vruno said.

Studies that require long-term datasets will be particularly susceptible to interference because there is a greater chance that leaky satellites will pass over them during the data collection period.

"As some projects need to continuously record data over hours, days, months or years, even tiny interference signals can have a statistical impact on the results," Winkel said. "And perhaps the astronomer analyzing the data is not even aware of it."

Intended downlinks, which are sent in multiple frequencies over 1 GHz, are also extremely bright and often appear in tandem with UEMR, exaggerating these effects.

As the problem gets worse, certain frequencies will become increasingly hard to study.

"Some radio bands could be completely wiped out," Di Vruno said. "And if [these] science cases are just not possible anymore, it would mean that we are effectively closing 'windows' to observe our universe."

Approaching the inflection point

As of May, there are 11,700 active satellites orbiting Earth. More than 7,300 of those (over 60%) are Starlink satellites, which have all been launched since 2019, according to Jonathan McDowell, an astronomer at the Harvard-Smithsonian Center for Astrophysics who has been tracking satellite launches and reentries since 1986.

But this is just the beginning. Well over 1 million satellites have been proposed by various organizations across the globe. And, while most of these may never be launched, many experts agree that we could eventually have up to 100,000 private satellites in LEO, potentially by as early as 2050. (This will likely be the maximum number that can be sustained at once without satellites crashing into one another.)

Related: What goes up must come down: How megaconstellations like SpaceX's Starlink network pose a grave safety threat to us on Earth

If that maximum number is reached, there is "real possibility" that we could reach an inflection point, beyond which ground-based radio astronomy would become effectively impossible, Di Vruno said.

Not all radio frequencies will be impacted. However, the obscured wavelengths will likely be lost for good, and the unaffected frequencies are unlikely to yield as many meaningful discoveries, he added.

At this point, we would no longer be able to "observe faint signals far out into the universe," which would be "a serious problem," Fionagh Thomson, a research fellow at Durham University in England who specializes in space ethics and was not involved in the LOFAR research, told Live Science.

Some radio astronomy could also still be achievable from space on a smaller scale. For example, there are plans to build a radio telescope on the moon. However, this would be very expensive and would capture limited data compared with the current suite of radio telescopes on Earth.

But even if we avoid the "worst-case scenario," we risk severely limiting our astronomical capabilities unless satellite operators and researchers can find viable solutions to the problem, Thomson said.

Plugging leaks and cleaning data

Satellite operators can limit the impacts of their spacecraft on radio astronomy in a few ways.

For example, most intended downlink frequencies are kept separate from those used by radio astronomers. Some companies, including SpaceX, are also investigating the implementation of "boresight avoidance," in which the satellites temporarily halt signal sending as they pass over radio "quiet zones," or areas where radio telescopes are actively collecting data.

However, for astronomers, it is also imperative that these companies minimize UEMR. We know that this is possible because spacecraft from NASA and other space agencies produce much less accidental radiation than private satellites do.

But companies can only mitigate a satellite's UEMR before it is launched into space. Once in LEO, "they are hard to fix," Winkel said, so it is vital that they are tested for leaks before launch.

"If satellite operators care about the UEMR, we will be OK," Di Vruno said. "It will be more difficult to conduct radio astronomy than it is now, but we understand technology evolves and we will evolve with it."

Astronomers can also limit the impacts of radio pollution by removing the interfering signals from their datasets. However, this "cleaning" may cause astronomers to miss crucial data that is masked by the interference.

"The amount of data you have to discard or the effort that you need to put in to somehow clean the data also increases the more interference there is," Winkel said. One way around this is to collect more data so that there is more left once you've cleaned it, but this also makes it much more expensive to do research, he added.

By working together, satellite operators and radio astronomers can solve the radio pollution issue without any external help, Thomson said. "But inevitably, satellite operators and the radio astronomy community have different goals, priorities and budgets, and finding workable solutions is no easy feat."

Laying down the law

Because private companies and scientists have different priorities, the most effective solution to the problem is to impose strict limits on the amount of UEMR that private spacecraft can give off, experts told Live Science.

"We would, of course, be more relaxed if proper regulation was in place," Winkel said.

At present, specific radio frequencies, like those used by LOFAR, are protected on behalf of astronomers by the International Telecommunication Union (ITU) — a United Nations agency responsible for regulating global communications technologies. However, these regulations apply only to Earth-based sources of radio pollution, not to private spacecraft.

Most satellite operators do try to respect the ITU's guidelines when using intended downlinks — with limited degrees of success. But UEMR frequently overlaps with the protected wavelengths and remains perfectly legal.

Some experts also argue that the ITU's radio-quiet frequency bands are no longer wide enough to protect radio astronomy: "They were set in a different era and are arguably too narrow for modern radio astronomy," Thomson said.

The IAU's CPU is hoping to have strict new regulations in place by the end of the decade and hopes that a breakthrough can be achieved at the next World Radiocommunication Conference, in 2027, Di Vruno said. Therefore, it is important for researchers to closely monitor the radiation leaking from private satellite constellations over the next few years, so that any new rules can have an effective and long-standing impact, he added.

However, even stricter guidelines might not be enough if organizations don't respect them.

"There is an assumption that imposing laws will fix complex problems," Thomson said. "But not all viable solutions involve implementing binding legislation," she warned.

In Birka, Sweden, there is a roughly 1,000-year-old Viking burial teeming with lethal weapons — a sword, an ax-head, spears, knives, shields and a quiver of arrows — as well as riding equipment and the skeletons of two warhorses. Nearly 150 years ago, when the grave was unearthed, archaeologists assumed they were looking at the burial of a male warrior. But a 2017 DNA analysis of the burial's skeletal remains revealed the individual was female.

Skeptics scrambled to explain away the evidence, said Charlotte Hedenstierna-Jonson, an archaeologist at Uppsala University in Sweden and first author of the 2017 study.

Even now, despite further studies strengthening the case for the Birka individual's martial profession, some archaeologists still insist she wasn't a warrior.

The Birka controversy highlights the fraught archaeological debate about the existence of Viking women warriors. Viking mythology and lore are filled with tales of women who lived for battle and engaged in violence, but whether these stories reflect real life is unsettled.

Across Scandinavia, at least a few dozen women from the Viking Age (A.D. 793 to 1066) were buried with war-grade weapons. Collectively, these burials paint a picture that clashes violently with the hypermasculine image of the bearded, burly Viking warrior that has dominated the popular imagination for centuries. And it's possible that, due to gendered assumptions, archaeologists may be systematically undercounting the number of Viking women buried with weapons.

The finds hint at a nuanced picture of Viking society — one where most warriors were men but a person's class and profession had the biggest impact on who went to war.

"Women can be as strong, as skilled, as fast as men," said Leszek Gardeła, an archaeologist at Ludwig Maximilian University of Munich and author of "Women and Weapons in the Viking World: Amazons of the North" (Casemate, 2021). "There is nothing in the biology there that would prevent them from being warriors."

Still, the poor preservation of Scandinavian graves, the enigmatic nature of Viking burials and the lack of historical texts leaves the meaning of many female burials up for debate. And even if women warriors existed, their significance in the broader Viking culture is unclear, Ole Kastholm, a prehistoric archaeologist and senior researcher at Roskilde Museum in Denmark, told Live Science.

"It's an area where we can't find a secure answer," he said.

Related: What's the farthest place the Vikings reached?

What the burials hold

Across Sweden, Norway, Denmark and Iceland, there are thousands of known Viking burials typically thought to be of male warriors. In contrast, we know of around 30 graves in which women were buried with obvious martial equipment such as spearheads and shields; of these roughly 30, only three have swords.

Of the known Viking Age burials, "statistically speaking, there would be less than 1%" of women buried with weapons among graves of men buried with weapons, Gardeła told Live Science.

But there are many more female burials that included other gear, such as shield bosses (a round protective metal piece at the shield's center), or possible weapons, such as arrowheads and ax-heads. Interpreting the latter burials is especially challenging because axes and arrowheads were used in battle, but they were also tools for hunting and farmwork.

But one of the main reasons the female-warrior question is so controversial is that many Viking burials aren't in great condition.

The Birka burial is one example. In 1878, workers used dynamite to blast open the grave, damaging it in the process. Untrained locals then helped excavate the grave. This poor excavation work has given naysayers room to argue that the chamber once held a double burial with a man.

According to the skeptics, "a woman would never be strong enough to use those weapons" — an argument that was ridiculous to Hedenstierna-Jonson, who had actually handled them.

"There were all these opinions rather than scientific facts," Hedenstierna-Jonson told Live Science.

Yet modern-era damage isn't the biggest obstacle to analysis. In many cases, bones and cremated remains are partially or completely decayed before archaeologists get a peek, largely due to Scandinavia's acidic soil. "We need very good preservation of the skeletons before we can determine the sex" via DNA analysis or bone studies, Kastholm said. "So even though the Viking Age has been investigated for like 150 years or more, it has not been that easy" to assess these graves.

As a result, archaeologists often guessed the deceased's sex based on grave goods, such as mirrors, weaving tools and brooches, which archaeologists assumed were typically buried with females, and battle-related weapons, which archaeologists thought were typically buried with males. If a Viking Age sword was the only item recovered, for example, it was nearly always assumed to be a male grave.

So it's possible archaeologists may be systematically undercounting Viking women who were buried with weapons.

"We could have a lot more of these [female] graves than we know about," said Marianne Moen, head of the Department of Archaeology at the Museum of Cultural History at the University of Oslo, calling the situation a catch-22. "You excavate a grave in Norway, you find a sword and you go, 'Oh, it's a man.' And then, 'isn't it funny how all the swords are buried with men?'"

In some instances, if a burial had both male- and female-associated artifacts, it is assumed, possibly incorrectly, that it was a double burial with a male and a female.

Even with that potential bias, there is strong evidence that some women were buried with war-related objects across Scandinavia. Norway has several of what have been nicknamed "shield-maiden" burials, after the women warriors of Scandinavian folklore. One is the Nordre Kjølen burial in Solør, which had a young adult — likely a female, based on a skeletal analysis — interred with a sword, an ax head, a spearhead, arrowheads, a shield boss, a horse skeleton and tools.

A second is the female boat burial from Aunvoll in Nord Trøndelag, in which a female was interred with a sword, eight gaming pieces, a sickle, a spearhead, shears, a knife and tools. The Klinta burial in Öland, Sweden has the cremated remains of what are thought to be an elite woman with valuable metal artifacts, including an ax-head, knives and an iron staff, causing some to wonder if she was a völva, or a Viking Age sorceress.

And although they're not buried with sharp weapons, "there's quite a few female burials on the west coast of Norway that have shield bosses and nobody likes to talk about them," Moen told Live Science.

Difficult to interpret

Still, many archaeologists struggle to make sense of these graves because the Vikings didn't have a consistent way of dealing with the dead.

"When we look at the Viking Age burials as a whole, they are weird and there's a great variation," Kastholm said. For instance, one had only a foot in it. Another was a triple burial of "a woman, then another woman buried some years later and then a half man buried later."

These mysterious, inconsistent burials make it hard to make straightforward conclusions.

Take the example of a grave discovered in Gerdrup, Denmark, in 1981, Kastholm said. A woman was buried with a spear, with large stones on her body. Her adult son, who had bound ankles and may have been hanged, was also in the grave.

The spear could be a sign the woman was a warrior. But that's not how Kastholm interprets the grave. Instead, he thinks that the son was hanged in devotion to Odin, the stones represent the woman's high status, and the valuable "spear was thrust into the bottom of the grave in a concluding ritual that dedicated the dead to Odin," he co-wrote in a 2021 study. This would have been a form of complex "mortuary theater," a play of sorts that would have been enacted at the grave site, which research suggests may have been common.

As for the Birka burial, Kastholm doesn't dispute that the deceased was biologically female and that she was buried with many weapons. "I'm totally convinced by that," he said. "If that means she was a warrior, I'm not convinced there. But that would go for male graves as well."

What historical texts tell us

To put the burials of women with weapons into context, archaeologists have looked at historical texts.

The Vikings left behind only a few thousand runic inscriptions. So most descriptions of warlike women and "shield maidens" come from semihistorical works written during the post-Viking medieval period. For instance, in "Gesta Danorum," a semifictional history of Denmark by Saxo Grammaticus (who lived circa 1150 to 1220), the warrior woman Lagertha travels with a group of women dressed as men, marries a Viking king who later divorces her, and still fights with him in a pivotal battle.

And some sagas, such as The Saga of Hervör and Heidrek, describe Norse women taking up arms to help protect family property, according to a 1986 analysis. Only men could inherit property, so if a man had only daughters, one was sometimes compelled to step into the role of a warrior as a "functional son" who could protect the family's interests, according to the study.

The Icelandic sagas, written by people who were likely the Viking's descendants in the 13th and 14th centuries, include stories about "women leading troops and engaging in acts of violence," Moen wrote in a 2021 article.

But are these stories evidence that Viking women were warriors in real life? Or did some stories have other mythical or mystical significance?

Some evidence points toward the latter. Sagas in which women wield weapons like axes often have magical overtones. In the Old Norse Ljósvetninga saga, for instance, a cross-dressing Norse sorceress strikes the water with an ax to see into the future. Axes are frequently associated with magic in folk traditions from Scandinavia, Finland and Central Europe, Gardeła noted in a 2021 article.

Gender wasn't destiny

After the Viking Age, the stereotype of the burly and ruthless male Viking warrior arose in the medieval sagas that detailed their exploits, and again in the late 19th century during Scandinavia and Iceland's National Romantic period. But it's possible that Viking Age society was "less governed by binary gendered ideals and more by fluid social obligations," Moen wrote in the 2021 article.

This would mean there wasn't a simple male-female dichotomy in who did what. This is seen in Viking Age grave goods. For instance, at Viking Age cemeteries in Vestfold, Norway, Moen found that although weapons were more common in male graves, they were also found in female burials. Likewise, while jewelry was more pronounced in female graves, 40% of male graves also had them, "hardly a negligible proportion," she wrote in the article.

Given how much violence permeated Viking society, "it would be naïve to think that only one half of the population was invested in it," she wrote in the article.

But people should not see this as female warriors filling a "man's role," Kastholm said. Rather, "warrior" was probably a profession, like modern-day firefighting, in which most were male but some were female.

Even among Viking Age men, being a "warrior" meant different things. Farmworkers, fishers and other peasants may have fought occasionally. But for the most part, the warriors were the social elite.

"Your biological sex was a factor [in your profession], but it was not the main factor," Hedenstierna-Jonson said. "The main factor was your role and your position and your family."

Still, people should be cautious in using information about these burials to infer how gender was perceived in Viking society, Moen said.

"I don't think it even necessarily indicates any kind of gender equality," she said. "What I do think is that you have much evidence women could be warriors and were warriors at certain times and in certain conditions."

"Occam's razor"

Moen splits archaeologists into three groups: those who think the burials clearly show that female warriors existed; people who say, "Yes, obviously women could be buried with weapons, but we need to question what it means"; and naysayers who think there's no way women actually used the weapons they were buried with. "They find it really quite troubling, and they go to very long lengths of explaining it away," she said.

To Moen, the evidence of female Viking warriors is right in front of us.

"Occam's razor, you know — the simplest explanation is usually the best," she said. "If you find a woman with a sword, then you should interpret it the same as you would a man with a sword."

In the end, Kastholm thinks "there will always be a lot of debate. And that debate is more about our time" and our modern-day attachment to gendered stereotypes about the Vikings than it is about the archaeological evidence, he said.

"Of course there were warriors in the Viking Age, and I'm pretty sure that some of them were female," Kastholm said. Yes, many graves are tricky to pin down, but at least a few have an impressive number of hard-core weapons buried with them.

"If it was a man," he said, "we would say 'that's a warrior grave."

Editor's Note: In this article, we are referring to biological sex, as it's impossible to know the gender of these deceased individuals.

Viking quiz: How much do you know about these seaborne raiders, traders and explorers?

A mysterious patch of water in the North Atlantic has baffled scientists for decades. Located to the southeast of Greenland, this blob of seawater was colder between 1901 and 2021 than during the late 1800s, even as the seas around it became ever warmer.

Some scientists linked this "warming hole" to an inflow of frigid Arctic meltwater; others blamed pollution from shipping, which can bounce the sun's rays back into space. But a growing body of evidence suggests the hole has a more sinister origin — one whose impacts, if unleashed, would reach much farther than a small region of the North Atlantic.

The warming hole might be a sign that Atlantic Ocean currents are slowing down. This network of currents regulates Earth's climate, moving heat from the tropics to the Northern Hemisphere. The currents, which include the Gulf Stream, form a huge and seemingly permanent loop known as the Atlantic Meridional Overturning Circulation (AMOC).

The AMOC has weakened before, ancient sediments reveal, and some past changes in circulation have been extremely abrupt. But human societies will struggle to adapt to the rapid rate of change that many researchers predict over the coming century, said Robert Marsh, a professor of oceanography and climate at the University of Southampton in the U.K. Weaker currents could unleash freezing weather in parts of Europe, exacerbate sea level rise along the U.S. East Coast, and trigger droughts around the equator in ways that scientists are racing to predict.

Whether the AMOC has already weakened and whether the currents could collapse completely are contentious questions. But what is clear is that "we're interrupting the system," Marsh told Live Science.

A giant heater

The AMOC is part of Earth's biggest conveyor belt, the thermohaline circulation, which pushes water around the world's oceans. Waters flowing north from the tip of South Africa ride on the surface of the ocean, absorbing heat from the atmosphere as they glide through the tropics and subtropics. They release this heat into the North Atlantic near Greenland, leading to a warming effect that is particularly strong in Northwest Europe.

"Right now, the Northern Hemisphere is warmer than it would be" if the currents that make up the AMOC were weaker, Maya Ben-Yami, an early-stage researcher who specializes in climate tipping points at the Technical University of Munich, told Live Science. For instance, the AMOC makes air temperatures in countries like Norway 18 to 27 degrees Fahrenheit (10 to 15 degrees Celsius) warmer, on average, than they would be otherwise.

Related: A long-lost ice sheet could predict the future of New York City — one in which Lower Manhattan and Coney Island are 'perpetually submerged'

To travel back south, surface waters in the North Atlantic must plunge down the water column and merge with bottom currents. Climate scientists say this sinking step is where the circulation is slowing down, because surface waters have to be very salty, and therefore very dense, to sink. But growing rivers of meltwater from the Greenland Ice Sheet and the Arctic are diluting the salt concentrations of surface waters, preventing them from plunging to the seafloor.

The warming hole that worries scientists is in a spot where the AMOC releases heat. Ocean temperatures there are lower than researchers would expect to see, given current global temperatures, if the AMOC were still going strong. That suggests heat transport from south to north has already declined.

And with climate change cooking the planet, there is a risk that more fresh water will pour into the North Atlantic, triggering a feedback loop that would accelerate the weakening of the AMOC. As ocean currents slow, the amount of salt reaching the North Atlantic may decrease, making it harder for already-diluted surface waters to sink.

"When you have a strong AMOC, it draws in salty water," David Thornalley, an ocean and climate scientist at University College London, told Live Science. "If you start to weaken the AMOC, less salty water gets drawn into the North Atlantic, and that then weakens the AMOC, and then it's a runaway process."

Defining "collapse"

The warming hole is one piece of evidence suggesting the AMOC has already weakened, but not everyone agrees.

The problem is that direct measurements of the AMOC's strength only go back to 2004, so researchers rely on indirect "fingerprints" — climate parameters they think are linked to the AMOC — to identify long-term trends in Atlantic circulation.

The trouble with fingerprints is that they give variable — and sometimes contradictory — results. For instance, a 2018 study suggested weakening, while research published in January suggests Atlantic currents have not declined over the past 60 years, though some scientists criticized the January study's data.

Regardless of whether Atlantic currents have already slowed, most scientists think they will in the future.

Should the circulation weaken significantly, "some pretty major changes are going to happen," Thornalley said, "and they're going to happen gradually over the 21st century."

For one, dwindling ocean currents will carry less heat to countries like Norway, Sweden and the U.K., leading to cooler temperatures in these regions. The amount of cooling will depend on how much strength the AMOC loses.

Climate models generate varying predictions, depending on what assumptions are baked into them, how sensitive they are to certain climate factors, and the data researchers feed the models. Some show a 10% reduction in the AMOC's strength, while others predict a decline closer to 50%, Marsh said.

"A 50% reduction would be far more substantial of an impact on our regional climate," Marsh said, adding that the resulting drop in temperature would disrupt industries that depend on the weather, such as farming. By comparison, "A 10% weakening of the AMOC would be noticeable, but manageable," he said. "It probably would make our weather a little less mild in the wintertime, in particular."

A complete collapse is unlikely, a study published in February found. Even if surface waters stop plunging in the North Atlantic, there are relatively constant factors, such as the wind, that drive the movement of water, Marsh said.

"A collapse, which is some hypothetical 100% failure of the [AMOC], just doesn't feel plausible to me," Marsh said. But the AMOC doesn't need to collapse to throw human lives and systems into chaos, he added.

"Really wild things"

Cooling in the North Atlantic and Northwest Europe could offset some of the warming from climate change — but that may not be a good thing, Marsh said. A drop in temperatures may alter atmospheric processes that drive storms and other weather events in these regions, he said.

Cooling in the North Atlantic won't directly make the U.S. East Coast colder, Marsh said. That's because weather in the Northeast is dictated more by the Arctic polar vortex, a ring of cold wind that circles the North Pole and traps frigid air near the Arctic. If this band becomes deformed or expands, which it sometimes does in the winter, chilly air escapes and sweeps over the Midwest and as far south as the Gulf.

But a weaker AMOC could alter the Arctic polar vortex, which would, in turn, affect the Northeast, Marsh said. The vortex is vulnerable to changes in climate, and recent evidence suggests the structure is already behaving strangely.

Direct impacts on the East Coast and Northwest Europe could also result from sea level rise. Sea levels in parts of the North Atlantic are currently 2.3 feet (70 centimeters) lower than they would be if the AMOC were weaker, because strong currents create uneven sea levels. If the AMOC weakens, water may redistribute itself and add to the sea level rise caused directly by global warming.

As the Northern Hemisphere cools, tropical and subtropical regions could also be thrown into chaos. Studies indicate that a colder North Atlantic would alter Earth's energy balance and trigger a southward shift of the Intertropical Convergence Zone (ITCZ), a band of clouds that encircles the globe near the equator.

The moisture-laden ITCZ bobs up and down depending on Earth's position relative to the sun, bringing heavy rainfall and storms. It is responsible for monsoons — weather patterns with distinct wet and dry seasons — in South America, West Africa, India and East Asia.

Should the ITCZ shift southward, research suggests it would throw monsoons out of whack for at least 100 years. "Right now, we have these regions that are used to getting this very intense rainfall in their wet seasons," said Ben-Yami, the lead author of this research. Ecosystems are adapted to these downpours, and people depend on rainfall for agriculture, she said.

A southward shift of the ITCZ would lead to shorter and drier wet seasons in places like West Africa. These effects could persist for decades and severely disrupt life around the equator, Ben-Yami said. "Farmers somewhere in sub-Saharan Africa, they need that rainfall," she said.

Shifts in rainfall would affect the Amazon rainforest, too. The effects may differ between the Northern and Southern parts, because the rainforest is so big that the southern part would still receive significant rainfall even if the ITCZ were to shift southward, Ben-Yami said.

But the northern Amazon rainforest — including parts of Colombia, Peru, Ecuador, Venezuela, Suriname, Guyana, French Guiana and Brazil — would experience longer and more intense dry seasons, which would be devastating for the region, she said.

Combined, the effects of a weakened AMOC could cost hundreds of billions, if not trillions, of dollars, Thornalley said. "It's a pretty global shift in climate — it would be very, very expensive," Thornalley told Live Science.

What's more, North Atlantic sediments from the last ice age (120,000 to 11,500 years ago) suggest that abrupt shifts in climate linked to the AMOC have occurred in the past.

"We know the climate can do really wild things," Thornalley said. "We know that really, really, really well."

Making predictions

Researchers say the AMOC will weaken significantly before 2100, but pinpointing exactly when the biggest change will happen is tricky.

Because direct measurements go back only 20 years, scientists often extrapolate from historical data to estimate when Atlantic currents might reach a tipping point. Tipping points are thresholds in some of Earth's systems that, once exceeded, can flip these systems from one stable state into a profoundly different one.

A good analogy to explain tipping points is a chair, Ben-Yami said. A chair is stable when it is standing on four legs or lying down; any position between these two is unstable. A person sitting on the chair can tip it backward slightly, but past a certain threshold, the chair tumbles to the ground.

An attention-grabbing study in 2023 concluded that the AMOC could reach a tipping point as early as this year. The authors used sea surface temperatures in the North Atlantic subpolar gyre between 1870 and 2020 to calculate fluctuations in the AMOC's resilience and strength. Their model suggested Atlantic currents are growing less resilient and weaker, to the point of possibly tipping before mid-century.

In a 2024 paper, however, Ben-Yami and other experts argued that there is too much uncertainty in data from before 2004 to predict when the AMOC might tip. "Once you take all these uncertainties into account, you don't get a practical prediction," she said.

Not having robust predictions is perhaps worse than knowing when the system will fail, because the future is so unclear, Ben-Yami said. "We should be even more careful" in the face of this uncertainty, she cautioned.

While some researchers are looking to the past for answers, others are searching for early warning signals that hold clues about future changes in the AMOC. One of these signals is the amount of fresh water flowing into the Atlantic at a latitude of 34 degrees south, near the tip of South Africa. Models suggest fresh water transport there will hit a minimum about 25 years before the AMOC reaches a tipping point, meaning experts could accurately predict shifts in the AMOC before the biggest changes occur.

But knowing about this minimum won't help anyone prevent the AMOC from weakening significantly, because it will be too late to reverse the driver — human-caused climate change — by the time the minimum is reached, said E.J.V. (Emma) Smolders, a doctoral student who specializes in AMOC early warning signals at Utrecht University in the Netherlands.

"We're getting more and more worried," Smolders told Live Science, adding that her research group is trying to find other warning signals that will help scientists make predictions with more lead time to act.

Her group described one of these alternative signals in a study published to the preprint database arXiv in June 2024. The study, which has not been peer-reviewed yet, found that salinity in the Atlantic near the tip of South Africa may be a better indicator than fresh water of when the AMOC will weaken.

Based on patterns in this salt transport, Smolders and colleagues estimated that the AMOC will "collapse" around 2050, which agrees with previous, peer-reviewed research.

Looking for early warning signals is important, because "it could be that we need as much advanced knowledge as possible to prepare civilian, even military, assets" to deal with the consequences of AMOC weakening, Marsh said. These assets may be needed to shield infrastructure and communities from unpredictable weather, he said.

Ultimately, regardless of whether we know exactly when the AMOC will decline, the takeaway is the same, Marsh and others said: Stop global warming by slashing the amount of carbon dioxide emitted into the atmosphere.

The possibility that the AMOC could be approaching a tipping point "should be enough incentive, alongside knowing what the impact would be, for us to want to stop what we're doing," Thornalley said. "We might not be able to say exactly how long we've got left, but [the evidence] suggests we could have a tipping point, and we really want to avoid that."

Editor's note: This article was updated at 5:14 E.D.T on April 25 to correct an error in the introduction. The blob in the North Atlantic is cold, not warm.

This story is part of The 89 Percent Project, an initiative of the global journalism collaboration Covering Climate Now.

On Oct. 29, 2012, Superstorm Sandy unleashed a 14-foot (4 meters) wall of seawater into New York City. Regionally, the storm killed 147 people, inflicted $50 billion in damage, and opened a window into the future.

By mid-century, flooding like this could become routine.

While the Statue of Liberty may not be submerged, low-lying places like Ellis Island could be underwater. By 2100, sea-level rise alone could approach Sandy's 2012 high water mark. The NYC Panel on Climate Change (NPCC) predicts up to 2.5 feet (0.76 m) of local sea-level rise by the 2050s and 9.5 feet (3 m) by 2100.

However, new evidence from Earth's past suggests these dire scenarios may be likely — and perhaps even conservative predictions.

Around 2.6 million years ago, the planet descended into the Quaternary, a period of alternating glacial-interglacial cycles that occurred roughly every 100,000 years. These cycles, driven partly by the planet's eccentric wobbles, tweak the amount of solar radiation reaching Earth. During cold (glacial) cycles, ice sheets sprawled across Antarctica, Greenland and North America, locking away water and lowering sea levels. In warmer (interglacial) times, the ice retreated and oceans rose.

The key to understanding today's rising seas may lie in the Last Interglacial (129,000 to 116,000 years ago), when global temperatures peaked at 1.8 to 3.6 degrees Fahrenheit (1 to 2 degrees Celsius) above preindustrial levels. Considering we're currently 2.7 F (1.5 C) above preindustrial levels, the Last Interglacial could presage changes in the coming decades as human-caused warming continues.

Related: Atlantic ocean currents are weakening — and it could make the climate in some regions unrecognizable

For years, researchers believed North America's Laurentide Ice Sheet disappeared very early in the Last Interglacial. But new evidence suggests the Laurentide lingered thousands of years after that. If true, this fundamentally alters our understanding of where and how quickly ice melted back then.

New models suggest that if the Laurentide persisted, Antarctica likely melted more — and more rapidly — than previously thought. And that has worrying implications for future sea levels.

"We all want to understand how small Antarctica's ice sheet became the last time Earth reached similar temperatures," Roger Creel, a geophysicist at Woods Hole Oceanographic Institution in Massachusetts who studies sea-level change, told Live Science. "Sea-level projections may be too low if Antarctic melt was underestimated."

Sea-level-rise projections

Global seas have risen a little less than 1 foot (0.25 m) since 1880. With the rate doubling since 2006, another quarter meter is all but certain by 2100. But that's likely an underestimate, as scientists expect accelerating polar melt.

The problem is pinning down how much the poles will melt.

Projecting sea levels is similar to forecasting a storm's path: We are trying to predict the future from past data. Unlike with a modern storm, however, very few of our sea-level records are definitive.

Direct measurements — taken from tide gauges and satellite data — capture just a blink of geologic time. The Liverpool tide gauge, one of the longest-running records, spans just 300 years, or 0.01% of the Quaternary. Predicting the future based only on direct measurements, then, is a bit like guessing a novel's ending from just the first few chapters.

Further back in time, the record gets murkier. Ancient glaciers obliterated much of the physical evidence that could help explain past climate changes. Where direct evidence survives — say, in ice cores or rubble piles — dating can be challenging. What's left is a patchwork of clues — fossils, sediments and ancient shorelines — that scientists piece together like detectives.

Patchwork of evidence

One person deciphering these ancient clues is Andrea Dutton, a University of Wisconsin-Madison geologist.

Far from Antarctica, jagged limestone scarps rise from the turquoise waters of the Caribbean. Beginning about 129,000 years ago, ice melted, seas swelled, and reefs grew to chase the sun. These reefs preserve evidence of sea-level fluctuations like mosquitoes in amber. Receding waters marooned these fossilized reefs on land, leaving them for researchers like Dutton to study.

Dutton has spent years tracking coral growth-and-death cycles to untangle how seas rose and fell during the Last Interglacial. Researchers still debate how high or fast seas rose, with estimates ranging from 6.5 to 33 feet (2 to 10 m) above present levels.

Dutton's work has taken her across the globe, and what she's found is perplexing. In the Caribbean, for example, sea levels peaked multiple times, whereas Chilean shores show no such patterns.

The public generally imagines uniform sea-level rise — a meter in Vancouver, a meter in Sydney. But that uniform rise "is a figment of our imagination," Dutton told Live Science.

In fact, deciphering ancient sea levels is maddeningly complex. That's partly because land surfaces aren't static, either. For instance, Dutton recalled researchers cataloging fossil corals in Papua New Guinea, only to find their study site uplifted by a sudden earthquake, which jumbled the geological record of historic sea levels. Aside from shifting coastlines, gravitational forces can distribute water unequally across the planet.

"That's why it's so important to look at many sites," Dutton said. "They all have different stories, yet clearly one thing must have happened in terms of global sea level."

Researchers initially attributed mysterious sea-level shifts — multiple peaks in some regions and falls in others — mainly to the melting and regrowing of Greenland and Antarctica during the Last Interglacial. But the idea of ice sheets regrowing when temperatures remained high never sat right with some. When evidence emerged from deep in the North Atlantic showing the Laurentide persisted, researchers needed a new model.

The lingering Laurentide

Creel set out to explain these mysterious sea-level patterns as part of his doctoral dissertation work under Jacqueline Austermann at Columbia's Lamont-Doherty Earth Observatory. Perched on rock once polished by the Laurentide Ice Sheet, the campus overlooks the Hudson River from a safe elevation above encroaching seas.

The Laurentide remains central to the big question: How high and fast will seas rise again?

Creel frames it like an accountant. Scientists typically budget interglacial sea levels from known suspects: ocean thermal expansion, Greenland, Antarctica and mountain glaciers. But the Laurentide adds a new line to the ledger.

For Creel and Austermann, timing is everything. "Antarctica is probably more vulnerable than we thought," Creel said. "If the Laurentide lasted longer, it may have masked how much Antarctica melted."

Once sprawling across 5 million square miles (13 million square kilometers) with ice up to 10,000 feet (3,000 m) thick, the Laurentide carved the Hudson River Valley and Great Lakes, growing and shrinking through ice ages. Today, evidence of it lingers in overgrown rubble piles and other topographic oddities. "The Laurentide is gone but not forgotten," Creel said.

For years, some scientists suspected a long-lived Laurentide, but hard evidence was scarce. That changed in 2022, when researchers studying North Atlantic sediment cores found signs of an enormous ice dam collapse 125,000 years ago — implying large amounts of North American ice.

A month later, a lead isotope study revealed the Laurentide lasted until around 122,000 years ago — well after temperatures rose during the Last Interglacial. And a 2023 study suggested this pattern wasn't unusual; sediments dropped from icebergs indicated the Laurentide melted only briefly and infrequently across multiple Quaternary interglacials.

To balance the books, Creel and Austermann fit together scattered and seemingly contradictory pieces of evidence — ice cores, fossils, sediments and geophysical properties — into a cohesive narrative.

A key element, they found in a study published late last year, was a process called isostatic adjustment. Ice sheets press down on the land; when they melt, the ground rebounds, altering Earth's rotation and gravity.

"It's like sitting on a cheap mattress," said Natasha Barlow, a paleoceanographer at the University of Leeds in the U.K. who was not involved in the study. "It sinks under your weight, and the other end rises — just like how ice compresses the Earth. When it melts, the 'mattress' rebounds, affecting sea levels differently depending on where you are."

Researchers already knew this process occurred, but the revised model's implications are striking: Antarctica experienced "rapid and drastic" ice loss early in the Last Interglacial — outpacing Greenland's melt — which caused the Antarctic crust to rise, sloshing water north. That raised sea levels in the Caribbean and other regions but caused them to fall near Antarctica. The process continued, lifting and dropping local seas differently in the global North versus the South.

"It was like a competition between the poles," Dutton said.

The model connected many puzzling data points, including the persistent Laurentide Ice Sheet and the confusing sea-level patterns of the Last Interglacial, Barlow said.

These findings help explain ancient sea-level puzzles, but they also hint at what's coming. If Antarctica melts faster than expected, places like New York will bear the brunt. While global seas may rise 6 feet (2 m) by 2100, New York could see a 10-foot (3 m) rise.

"For many years, we've said it's all about Greenland and Antarctica," Barlow told Live Science. "But the evidence over the last five years says, 'Hold on a second. Maybe the Laurentide persisted longer than we thought.' This is the first study seriously considering its influence at this level. If we're going to calibrate our models for the future, we need to think about this."

All eyes on West Antarctica

Because most of Antarctica's ice lurks below sea level, it is more vulnerable to melt than Greenland's land-based ice — precariously so in West Antarctica.

In 1978, glaciologist John Mercer warned that rising carbon dioxide levels could trigger the collapse of the West Antarctic Ice Sheet, causing 15 feet (5 m) of sea-level rise and the "submergence of low-lying areas." His claims, which initially drew accusations of alarmism, now appear eerily prescient.

Whether seas rise slowly or exponentially depends on carbon dioxide emissions and ice sheet behavior.

In the past decade, researchers have coalesced around the idea that marine ice sheet instability (MISI) may lead to considerable West Antarctic melt. In MISI, warm water melts ice along Antarctica’s coast, causing the grounding line — the ice's contact with the seafloor — to retreat. As the grounding line retreats inland, ice flows faster and more icebergs break into the ocean, accelerating ice loss in a self-reinforcing cycle. MISI, other ice retreat processes, and high-emission scenarios underpin the Intergovernmental Panel on Climate Change's (IPCC) worst-case predictions.

Mathieu Morlighem, a Dartmouth glaciologist who studies ice sheet physics, believes MISI may be inevitable.

"In all our simulations, West Antarctica collapses," Morlighem told Live Science, referring to model projections over the next few centuries. West Antarctica alone could contribute over 10 feet to global sea levels.

Evidence from the Last Interglacial also supports the modern observations that West Antarctica is especially prone to melt. Sediment and ice core studies published in 2021 and 2023 suggest West Antarctic ice melted or thinned significantly during the last warm period. Another 2023 study found octopus populations around West Antarctica interbred during the Last Interglacial, which would have been unlikely with an intact ice sheet.

Combined with data about the Laurentide, the emerging picture suggests West Antarctica melted significantly in the Last Interglacial and may do so again over the next several decades.

From the past comes the future

If recent findings hold up and significant West Antarctic melt is imminent, even our gloomiest predictions may be optimistic.

In the IPCC's "high impact, low probability" 2100 scenario, sprawling and vibrant low-lying areas — including the Jersey Shore; Manhattan's East Village and Financial District; the Rockaways; Coney Island; Red Hook, Brooklyn; and parts of Staten Island — would be perpetually submerged.

New York already experiences "sunny day flooding," but storm surges would magnify the damage. The NPCC concluded that, by the 2080s, rising sea levels alone — without any changes in storm frequency or intensity — could make today's "once-in-a-century" coastal floods happen twice as often, or even 10 to 15 times more frequently.

"I can't dunk. But if I raise the court a foot, my chances are better. Boost sea level, and you're looking at the potential for much greater flooding risk," Daniel Bader, a climate scientist at Lamont-Doherty and an NPCC member, told Live Science.

And the new insights from the Last Interglacial could lend weight to this scenario, which was once dismissed as extreme.

Still, several experts cautioned against drawing too many lessons from the past. Ancient clues are scattered, dates are imprecise, and the drivers of warming were vastly different then. Yet some experts told Live Science they believe the IPCC's worst-case models are, if anything, too conservative.

Unlike warm periods driven by natural processes, today's warming feels more dire, Dutton said. "Today we're forcing both [natural and human-induced warming] simultaneously," she explained. "Greenhouse gas concentrations are heating both poles at the same time." Warming during the Last Interglacial, by contrast, was likely focused more on the Northern Hemisphere.

For Klaus Jacob, a Columbia University geophysicist and climate mitigation expert, whether the worst-case scenarios are likely is beside the point. The possibility demands action: "Would you board a plane with a 10% likelihood of crashing?" Jacob said.

Jacob envisioned three possible responses: protection, with barriers like those that defend the Netherlands but failed catastrophically in New Orleans during Hurricane Katrina; adaptation, such as using stilts, abandoning ground floors, and embracing Venetian-style boat transport; and relocation.

New York only began earnest reflections after Superstorm Sandy but remains mired in bureaucratic inertia, Jacob said. Coastal defenses are stopgaps, he argued; the real solution is managed retreat. In other words: "Get the hell out."

That means moving to higher elevations, and potentially abandoning entire low-lying regions along the U.S. East Coast, Gulf Coast and swaths of Asia. And if the retreat isn't managed?

"It's chaos," Jacob told Live Science. "Chaotic retreat."

This story is part of The 89 Percent Project, an initiative of the global journalism collaboration Covering Climate Now.

You're holding two wrinkly human brains, each dripping in formaldehyde. Look at one and then the other. Can you tell which brain is female and which is male?

You can't.

Humanity has been hunting for sex-based differences in the brain since at least the time of the ancient Greeks, and it has largely been an exercise in futility. That's partly because human brains do not come in two distinct forms, said Dr. Armin Raznahan, chief of the National Institute of Mental Health's Section on Developmental Neurogenomics.

"I'm not aware of any measure you can make of the human brain where the male and female distributions don't overlap," Raznahan told Live Science.

But the question of how male and female brains differ may still matter, because brain diseases and psychiatric disorders manifest differently between the sexes. Disentangling how much of that difference is rooted in biology versus the environment could lead to better treatments, experts argue.

There are many different disorders of the brain — psychiatric and neurologic diseases — that occur with different prevalence and are expressed in different ways between sexes, said Dr. Yvonne Lui, a clinician-scientist and vice chair of research in NYU Langone's Department of Radiology. "Trying to understand baseline differences can help us better understand how diseases manifest."

Now, thanks in part to artificial intelligence (AI), scientists are starting to reliably distinguish male and female brains using subtle differences in their cellular structures and in neural circuits that play a role in a wide range of cognitive tasks, from visual perception to movement to emotional regulation. Other studies point to sex-based differences in human brain structure that may be present from birth, and still other, lab-based research in animals points to sex-based differences in how brain cells fire at a molecular level.

What's still completely unclear is to what extent these differences matter. Do they change how people's brains function or how susceptible they are to disease? Should they dictate which treatments doctors offer to each patient? Even as scientists pinpoint subtle brain differences between females and males, their research inevitably runs up against tricky questions of how sex, gender and culture interplay to sculpt human cognition.

Right now, it's impossible to answer these big questions. But ongoing and future research — focused on lab animals, human chromosomes and brain development, and subjects followed from youth through adulthood — could start to reveal how these sex-based differences concretely affect cognition, and ultimately, the development of diseases of the brain.

Why study sex-based brain differences?

Historically, scientists used purported brain differences to make sweeping statements about how men and women think and behave and to justify sexist beliefs that women were innately less intelligent and less capable than men.

While that early research has been discredited, modern studies still find cognitive differences between men and women — at least on average. For example, men reportedly perform better on tests of spatial ability, while women are better at interpreting the facial expressions of others. But men and women are raised and treated very differently in society, so what's at the root of these differences? Is it nature or nurture, or both?

"It's actually incredibly difficult in humans to … causally distinguish how much of a sex difference is societally or environmentally driven," Raznahan said. "We have all of these assumptions and biases that sort of slip into our heads through the back door without us realizing."

Given the dubious history of studying sex differences in the brain, and the logistical difficulty of doing it the right way, one might wonder why scientists bother. For many, it's because neurological diseases and psychiatric conditions seem to play out differently in males and females, and both biological and environmental factors could explain why that is.

Data suggest women experience higher rates of depression and migraine than men do, while men have higher rates of schizophrenia and autism. About twice the number of men develop Parkinson's disease than women do, but women with the condition tend to have faster-progressing disease. All these data come from studies that don't necessarily distinguish sex from gender — "sex" describes biology, while "gender" reflects self-identity, as well as societal roles and pressures. Lumping the two concepts together muddies our understanding of why a given difference exists.

For instance, pubescent girls are more likely to experience depression than boys are, which may be related to how their maturing brains handle stress or the possibility that they encounter more stressful events than boys do at that age. Conversely, do boys' brains make them resilient against depression, or are they actually going underdiagnosed due to social stigma? The answers to these questions point to different solutions.

Large-scale structures, negligible differences

Thanks to brain-scanning techniques like MRI, scientists have found subtle sex differences in the size, shape and thickness of various brain structures, as well as differences in networks that link different parts of the brain.

But these differences are small to negligible when you account for the average size difference between males and females, argues Lise Eliot, a professor of neuroscience at the Rosalind Franklin University of Medicine and Science and author of "Pink Brain, Blue Brain" (Houghton Mifflin Harcourt, 2009).

Eliot and colleagues recently looked at about 30 years of studies, finding that, on average, male brains are 6% larger than female brains at birth and grow to be 11% larger by adulthood. This makes sense because average brain size scales along with average body size, and male bodies tend to be larger. But when you take this overall size difference into account, subtler structural differences between male and female brains shrink to the point of negligibility, the researchers concluded.

"There are maybe species-wide sex differences in the brain, but so far, they haven't been proven," Eliot told Live Science. "And so if they exist, they must be pretty small."

Nonetheless, some scientists have reported differences that they say don't scale with body size. Some examples came from a research group who'd crunched MRI data from over 40,000 adult brains scanned for the UK Biobank, a repository of medical data from 500,000 adults in the United Kingdom.

In that study, males had a larger thalamus, a relay station for sensory information. They also had a larger putamen, which helps control movement and forms part of a feedback loop that tells you whether a movement was well executed. Females, on average, had a larger left-side nucleus accumbens, part of the brain's reward center, and a bigger hippocampus, the storage site for short-term memories of facts and events that also helps transfer the information to long-term memory.

But neither this nor other studies have revealed a specific feature that reliably distinguishes a given male brain from a female brain, since the size ranges seen in each sex largely overlap, Raznahan and colleagues noted in a letter responding to that study.

For the few size differences that do exist, it's currently impossible to say whether they explain any differences in cognition linked to sex, or alternatively, whether they actually make males' and females' cognition more similar, the letter authors noted. Perhaps male and female brains operate slightly differently to reach the same output — to "counterbalance" differences in hormones or genetics that may affect brain function, they wrote.

"When we're just talking about describing a difference in a measurement, that's not saying anything about whether it's got any functional relevance at all," Raznahan emphasized.

AI finds subtle differences

While large-scale structural features might not distinguish male and female brains, AI is helping to uncover other, subtler features that may differentiate the two. Some of these differences appear on the level of the brain's microstructure, meaning its individual cells and components of those cells.

For instance, a study published in May 2024 used different AI models to analyze brain scans from 1,030 young adults ages 22 to 37 years old. The research primarily focused on white matter, the bundles of insulated wiring that run between neurons. "I believe ours is the first study to detect brain microstructural differences between sexes," said Lui, who co-authored the study.

The AI models analyzed differences in both local landmarks in the brain — such as the corpus callosum, which connects the brain's two halves — and the highways that connect distant cells. It also looked at differences in how the white matter was bundled together, as well as in how dense and well insulated those bundles were.

The algorithms accurately predicted the sex of the subject tied to a given scan 92% to 98% of the time. That remaining gap in accuracy likely comes down to the "huge amount of variance in humans," Lui said.

No single part of the brain could be used to make predictions; one model relied on 15 distinct regions of white matter. All models showed some consistencies, though, with the largest white matter structure that crosses the midline, the corpus callosum, standing out as key.

From birth

Lui and colleagues' study was not designed to address how an individual's upbringing or environment shapes the brain. Nor did it aim to disentangle biological differences in the brain from those rooted in gender.

Sex describes biological differences in anatomy, physiology, hormones and chromosomes. Sex traits are categorized as male or female, although some people's traits don't fit neatly in either category. Gender, on the other hand, is cultural. It encompasses how people identify and express themselves, as well as how they are treated and expected to behave by others. Genders include man and woman, as well as others, including those that fall under the umbrella term nonbinary or are unique to specific cultures, like the māhū of Hawai'i.

Historically, studies have conflated sex and gender. To tease these factors apart and see how each manifests in the brain, it would be helpful to follow people over time as their brains are developing — and new research is beginning to do just that.

For example, a 2024 study looked at average brain volume in over 500 newborns: Males' brains were 6% larger overall, even after accounting for differences in birth weight, and females had larger gray-to-white matter ratios. (Gray matter, the cell bodies of neurons, is primarily found in the outer layer of the brain, called the cortex.) That average difference in gray matter is also seen in adults, which makes sense given that larger brains need more white matter to relay signals between far-apart cells.

Statistically, these big-picture brain differences were more significant than differences seen in smaller structures. Females had larger corpus callosa, as well as more gray matter around the hippocampus and in a key emotion-processing hub called the left anterior cingulate gyrus (ACG). Males had more gray matter in parts of the temporal lobe involved in sensory processing, as well as in the subthalamic nucleus, key for movement control. But sex could only explain a fraction of the variance seen in these structures.

Some of these brain differences are "present from the earliest stage of postnatal life" and persist into adulthood, the authors noted. This applies mostly to the global differences, but also potentially to some of the smaller ones. For example, some studies — but not all — show that the left ACG is also larger in adult females, not only in babies.

Durable differences present from birth are likely sex-based. But differences that emerge or disappear in later life, like those in the hippocampus, may be influenced by the environment, or else reflect sex differences in development, including hormonal shifts in puberty.

Gender and sex

Studies like this can help tease apart the influence of sex and gender on the brain. At present, there's a "massive gap" in our understanding of how these factors shape the brain independently and in tandem, said Elvisha Dhamala, an assistant professor of psychiatry at the Feinstein Institutes for Medical Research in New York.

Dhamala and colleagues recently aimed to fill in that gap using data from the Adolescent Brain and Cognitive Development (ABCD) study, an enormous U.S.-based study of brain development and child health. They incorporated functional MRI (fMRI) scans from nearly 4,800 children; fMRI tracks blood flow in the brain to give an indirect measure of brain activity. Each child joined the study at age 9 or 10 and will be followed for 10 years, which will enable follow-up studies.

The fMRI scans highlighted linked brain areas, or networks that lit up as the children did different tasks, including memory tests that required them to recall several images. The children and their parents also answered questions about the kids' feelings about their genders and how they typically play and express themselves. "It's not anything clinical," Dhamala noted. "It's just an aspect of behavior that represents your gender."

These answers were used to generate "scores" for each child that the AI algorithm could use as data points.

The algorithm ultimately revealed two largely distinct brain networks tied to sex and gender. The brain differences most strongly tied to sex were found in networks responsible for processing visual stimuli and physical sensations, controlling movement, making decisions and regulating emotions. Differences tied to gender were more widely dispersed, involving connections within and between many areas in the cortex.

After pinpointing these networks, the researchers trained their AI algorithms to "predict" a child's sex or gender based on brain activity. They accurately determined most children's sexes, similar to the results of Lui's study. Gender proved trickier: With the children's questionnaire answers, the AI couldn't predict where they landed on a continuum of gender, whereas with the parents' answers, its predictive power exceeded chance but was still "much lower" than the predictions for sex, Dhamala said.

Nonetheless, the study highlighted an understudied idea: that gender sculpts the brain in ways that are distinct from sex, she said.

Interestingly, some tentative lines can be drawn between Lui's and Dhamala's AI-powered studies. They can't be directly compared, as the two studies used different types of analyses and focused on different features of the brain. But many of the physical white matter tracts flagged in the former study correspond with functional networks highlighted in the latter, Dhamala told Live Science.

As an example, the cingulum — a white-matter tract that encircles the corpus callosum — seemed key for making predictions in Lui's study. It also links together various networks flagged in Dhamala's study, including circuits involved in emotional processing. That hints that sex differences exist in both the physical anatomy of these networks and in their activation patterns, Dhamala said.

The future of the sex-difference field

Scientists have made some progress at teasing out sex differences in the brain, but to truly understand these distinctions, researchers will need to do more animal studies to allow for more experimental control, according to a 2020 paper co-authored by Raznahan.

Various studies in lab rats have already revealed differences in how males and females form connections between neurons, and how each sex processes fearful memories, for example.

Related: 'Let's just study males and keep it simple': How excluding female animals from research held neuroscience back, and could do so again

In humans, scientists can collect more brain data right at the time of birth, to pinpoint baseline differences that might exist before a child encounters any cultural influences, and then track the child over time, Raznahan and colleagues added.

Another option is to study human genes that are unique to either the X or Y chromosome. By looking at people with extra or missing sex chromosomes, for example, scientists have started to unravel how these genes either inflate or shrink brain structures, contributing to sex differences in size. Chromosomes may also raise or lower the risk of disorders — for instance, carrying an extra Y raises the likelihood that a person has autism, whereas an extra X does not. That may help to explain why males, who usually carry one X and one Y, have higher autism rates than females, who typically have two Xs.

Right now, the fate of such research is uncertain in the U.S.

Prompted by executive orders from the new presidential administration, the National Science Foundation has been combing through active research projects to see if they include words that might violate said orders, such as "woman," "female" and "gender," and the National Institutes of Health appeared to archive a long-standing policy requiring both male and female lab animals in studies.

"There's just a lot of uncertainty," Dhamala told Live Science. If the worst case scenario comes to pass, "removing that gender component, or making it harder to study sex differences, is going to push us backward rather than forward."

But if the field survives, future work could incorporate gender the way the ABCD study did, using questionnaires to generate composite scores, Dhamala said. As a start, scientists could at least ask study participants what gender they identify as, she added. Other experts agree.

By adopting these strategies, scientists could dramatically advance this research field that dates back to Aristotle. Their efforts could lend new talking points to the endless debate of nature versus nature. They could uncover meaningful sex differences that pave the way to better treatments for depression, Alzheimer's and more. Or they could highlight the ways members of the "opposite sex" are actually more alike than they are different.

There's a spa floating in the middle of Lake Erie. It has a sauna, a steam room and even a cubicle filled with snow. Upstairs, there are luxury lounges, a huge library, a curated art collection by notable artists, and a panoramic lecture theater with floor-to-ceiling windows. Passengers are busy dining, surrounded by sommeliers, in fine restaurants.

One deck below, there's a pristine, state-of-the-art laboratory full of high-tech equipment, and two multimillion-dollar submersibles can take passengers down 1,000 feet (300 meters). A team of scientists is sifting through water samples and analyzing them in real time, looking at the genetic fingerprints of plankton as it floats through the water.

The researchers on Viking's Octantis cruise ship are studying environmental DNA (eDNA) — bits of genetic material that float in the water, drift through the air, or linger in the soil. Every time a living creature passes through an environment, it sheds minuscule bits of its genetic material.

Scientists first noticed traces of this genetic material decades ago, but thanks to powerful sequencing techniques, they are now beginning to analyze eDNA to characterize food webs, reveal the locations of long-lost endangered species, and show if predators are lurking in areas where humans and wildlife are in conflict.

But the technique has one problem: It generates so much data that researchers struggle to analyze it all. Now, scientists are working to combine artificial intelligence (AI) with cutting-edge sequencing to rapidly identify changes in the types and numbers of organisms in a given ecosystem. Eventually, that information could provide a real-time view of how the planet operates — and allow us to adapt to ecological changes more quickly.

"AI's going to be able to pull out [information] in a way that our other techniques just don't have the capabilities to," said Zachary Gold, research lead of the Ocean Molecular Ecology program at the National Oceanic and Atmospheric Administration's (NOAA) Pacific Marine Environmental Laboratory. "Quicker, better, faster data allows us to do things we've never dreamt of before," he told Live Science.

A treasure trove of environmental data

The term "environmental DNA," or "eDNA," was coined in the 1980s in a study describing a technique for getting DNA from a soil sample. But it wasn't until the 2000s that fast and accurate DNA sequencing machines became widely available and affordable, making eDNA analysis practical.

Next-generation sequencing (NGS) now allows scientists to analyze DNA incredibly quickly — the entire human genome can now be sequenced in just one day. For eDNA, NGS means thousands of species can be identified from a single water sample. The sequencing technology is highly advanced, but the ability to analyze and draw meaningful conclusions from it requires a huge amount of computing power and could take years of scientists' time.

The physical samples can take anywhere from a couple of days to a month to sequence, then once the sequences come back, many gigabytes of data must be downloaded and "cleaned" — that is, checked by a computer for mistakes, duplicates or formatting issues. Only then can validated datasets be analyzed.

It's that next step where AI could be transformative.

"Researchers can spend months looking through that data to try to understand and identify what are the most interesting and more powerful stories and assets that are coming out of this data, but the AI could do it, you know, in seconds," Gold said.

An army of floating laboratories

Viking began studying eDNA in part because of the pandemic. The company initially required guests to take daily polymerase chain reaction (PCR) tests for COVID-19, but once that requirement was phased out, the equipment on board its ship Octantis was repurposed to allow for real-time testing of water samples. The cruise company teamed up with NOAA in 2020, and scientists joined Viking's expedition to the Great Lakes in 2022.

Now, scientists aboard this 673-foot-long (205 m) cruise ship analyze phytoplankton in the waters they pass through, providing a snapshot of the ecosystem each time the ship visits the same regions. Compared with traditional scientific research expeditions, which are expensive and irregular, tourism vessels save time and money — cruise ships are going on these voyages anyway — and the food is a lot better, the team said.

In their floating lab, researchers working with Viking now sequence phytoplankton. "They are the key to life on Earth," said Allison Cusick, a researcher at the Scripps Institution of Oceanography at the University of California, San Diego, who works in one of Viking's ship laboratories to study eDNA in remote locations like Antarctica. Phytoplankton are the foundation of most marine food webs, and they produce about half the planet's oxygen via photosynthesis. The differences among phytoplankton species is mind-blowing — the diversity between two types can be greater than that between a human and a fungus, Cusick said.

Changes in the type of plankton in the water are key indicators of biodiversity and ocean health — shifts can ricochet up the food web, with potentially devastating consequences.

Using eDNA analysis to uncover evolutionary relationships between species and the different evolutionary paths they took — for example, when one arose and when specific genes were introduced — could help scientists predict how climate change will affect different species, said Benoit Morin, a supercomputer engineer at IFREMER (the French National Institute for Ocean Science and Technology).

"By looking at the past, we can try to understand the future," Morin told Live Science.

An "Enigma project" for eDNA

To be really powerful, projects like the Viking-NOAA collaboration will need to integrate artificial intelligence into eDNA analysis.

Already, AI is being used to find potentially new species from large data sets from camera traps and automated monitoring systems. It's also being used to rediscover lost species, including the critically endangered De Winton's golden mole (Cryptochloris wintoni), which, until it was traced using eDNA, hadn't been seen for over 80 years.

But for these efforts to reach their full potential, AI techniques will need to be refined and integrated into eDNA analysis.

Once scientists have collected an eDNA sample, they analyze it via bar coding, which can either look for a single species or organism or identify multiple species at once. The barcode is a small series of unique DNA sequences that are used to identify an organism by comparing it to an online reference database.

Letizia Lamperti, a mathematical engineer at the École Pratique des Hautes Études (Practical School of Advanced Studies) in France, is developing a machine learning system to use such bar coding to reveal the health of a given environment, based on the type and number of organisms within a sample. That information, in turn, could point to potential fixes.

For example, if there was an increase in toxin-producing phytoplankton in a water sample, it may be possible to pin those changes to agricultural runoff that's feeding the phytoplankton, Cusick said.

In 2023, Lamperti and her colleagues published a study showing that neural networks — multilayered machine learning algorithms that mimic the way the human brain filters and processes information — do a better job than other statistical methods of grouping closely related organisms based on their eDNA. But just like facial recognition technology, AI will likely be better at detecting abundant species, for which there is a lot of "training" data, but less effective at spotting rarer organisms.

Several other recent studies point to the promising potential for AI in eDNA research. For instance, one study found that AI can identify 90% of unknown species in a sample, even when there aren't similar sequences from closely related organisms to use for comparison.

If AI can fulfill its potential, the shift in how we understand the environment would be monumental. Cusick likened it to Alan Turing's decryption of the Germans' Enigma code during World War II. "That's going to be transformative," she told Live Science.

AI could identify newfound species on an unparalleled scale. Evolutionary relationships could be determined in the blink of an eye. Monitoring and planning for environmental changes could be transformed. For instance, by rapidly analyzing eDNA samples, AI could alert swimmers in real time to the presence of brain-eating amoebas or sharks in waterways, or forecast events like harmful algal blooms before they threaten public health — similar to how we get weather alerts on our phones now.

In theory, then, resources could be redirected quickly to resolve issues before they become a problem.

This goal is achievable, Gold said, but how long it will take will depend on the resources funneled into developing the AI to do so.

A dictionary of species

At the moment, AI is missing something important: organized volumes of good data for spotting key patterns. These data need to be put in one place as a reference database, or a dictionary of species, based on their DNA.

"We need the database of reference to perform the species identification," Lamperti told Live Science. "The problem is that we don't have it."

To identify species, AI needs to learn the key signatures, or barcodes, of individual and closely related species by training on reams and reams of data. But biodiversity datasets are not in publicly available repositories, and they're not in curated, standardized formats that can be fed into trained, bespoke AI systems. "eDNA is not AI-ready," Gold said.

In the U.S., around 40,000 eDNA samples have been collected in the past decade alone, Gold estimated, but a lot of it isn't accessible. It could be "in somebody's attic or the supplemental methods of someone's scientific paper," he said.

To draw useful conclusions to help us protect and manage the environment, AI needs to learn from a baseline database that captures biodiversity in the environments we're interested in. That's a herculean effort. "It's millions of dollars; it's tons of people's time," Gold said.

Morin is currently working on this task, but it's a slow and resource-intensive process. He and his colleagues are building a genetic "dictionary" through the ATLASea project, which aims to sequence the genomes of 4,500 marine species. This information will be deposited in an open-access database for the scientific community. IFREMER is now working with data infrastructure company NetApp to classify the mass of information being collected.

With money to develop the datasets, an AI eDNA tool could be ready "really fast," Gold said. "I have no doubt that what we're doing is not technologically difficult. It's just we're not resourcing it. If we really wanted to do this and mobilize at a scale, I have no doubt by the next Olympics in Los Angeles [in 2028], we could have the tools and resources and network set up and [be] ready to do this."

If investment and resources continue at their current pace, Gold estimated it will be a "slow trickle" and we'll get there in around 15 years. But he's optimistic the timescale could be faster. "A lot of the stuff isn't hard; it's just taking the existing tools that are already out there," Gold said. "We've just got to point the bike in the right direction."

Editor's note: This article was updated at 4.25 a.m. ET on Feb. 25 to to clarify that Viking, not external agencies, required guests to take daily PCR tests during the COVID-19 pandemic.

In Science Spotlight, Live Science takes a deeper look at emerging science and gives you the perspective you need on these advances. Our stories highlight trends in different fields, how new research is changing old ideas, and how the picture of the world we live in is being transformed thanks to science.

When Alexander the Great died in 323 B.C., 13 years after building one of the world's largest empires, he left what would become one of the biggest unsolved mysteries in archaeology.

Researchers have hotly debated where Alexander the Great is buried, with theories ranging from his homeland in Macedonia (now Greece) to Egypt.

Like the Holy Grail or the Ark of the Covenant, the tomb has taken on a mythical status that has captivated many archaeologists over the years.

"As long as his tomb isn't found, he lives on in a quasi-mystical limbo, forever sparking new ideas and theories and controversies," Nicholas Saunders, a professor emeritus in the Department of Anthropology and Archaeology at the University of Bristol in the United Kingdom and author of the book "The Life of Alexander the Great" (Brighter Child, 2006), told Live Science in an email.

Hundreds have hunted for the tomb, to no avail.

"I remember in 1963 when I was a student at Alexandria University, there was a Greek man who was doing excavations near the [train] station," Zahi Hawass, an archaeologist, Egyptologist and Egypt's former minister of antiquities, told Live Science. "I used to go and see his excavations. He found nothing, but he stayed five years, searching."

But have archaeologists made any progress on finding the tomb? Calliope Limneos-Papakosta, director and founder of the Hellenic Research Institute of Alexandrian Civilization, who is currently excavating in Alexandria, Egypt, thinks she's close.

The tomb may be hidden beneath the streets of modern-day Alexandria, at the intersection of two ancient thoroughfares, she said. An intriguing Hellenistic statue and other signs of ancient quarters from the right period point to a very promising location.

"I am at the crossroads, and I believe that I have more possibilities to find the tomb than anybody else," Limneos-Papakosta told Live Science.

Why finding the tomb is so difficult

One reason the tomb still hasn't been found is because Alexander's body was moved, and so he was buried more than once.

In addition, historical records of the tomb are scarce and were often written hundreds of years after the events they describe. Each of those sources had a political agenda or bias, said Paul Cartledge, a professor emeritus of Greek culture at the University of Cambridge.

"So, you have to take everything with a large pinch of salt," Cartledge, author of "Alexander the Great" (Overlook Press, 2004), told Live Science.

Compounding the difficulty is that there are no clear images of the tomb that could guide archaeological searches, Saunders added.

Finally, the modern city of Alexandria was largely built over the original.

"Modern Alexandria, especially the downtown area, was built completely above ancient monuments," Hawass said. "The theater was found while they were building houses. And catacombs were also found."

To find the tomb, archaeologists would have to dig up places where people live and work today, which is challenging without strong evidence that something important lies beneath.

What we know about Alexander's death and burial

We know from historians like Plutarch that, at age 32, Alexander died sometime between the evening of June 10 and the morning of June 11, 323 B.C., in Babylon (modern-day Iraq). His cause of death is controversial, with some proposing he died of diseases such as typhus or malaria or from drinking poisoned wine, and others arguing he was assassinated. (A few people even think he was declared dead prematurely.)

Related: How did Alexander the Great die?

Regardless of what killed Alexander, his body was mummified. Ptolemy I Soter, a Macedonian Greek general who was Alexander's close friend and bodyguard, was instrumental in moving the body.

"His corpse was taken en route via the Middle East through what is now Iraq to what's now Syria," Cartledge said. From there, over two years, it would travel almost 1,000 miles (1,600 kilometers) to Egypt.

Ptolemy had an agenda: Alexander had left no clear political heir (although his wife Roxana was pregnant with his son and he may have had an illegitimate heir), and building a tomb for Alexander was a way to cement Ptolemy's position as the legitimate successor to the Macedonian king, Cartledge said.

Alexander's body was then shipped to the Valley of the Kings near Luxor, Egypt, where many of the New Kingdom pharaohs were buried, while the newly founded city of Alexandria was being built.

"In 305 B.C., 18 years after he died, Alexander's corpse was brought to Alexandria, by which time there's a royal court, a palace and a museum," Cartledge said. "Having Alexander buried there would have been the icing on the cake" for the nascent city.

A few decades later, accounts by Greek geographer Strabo suggest Ptolemy IV Philopator, the fourth pharaoh of Ptolemaic Egypt and the great-grandson of Ptolemy I, moved Alexander's body to its final documented resting place: Soma, Alexandria's great mausoleum, according to the book "Alexander's Tomb: The Two Thousand Year Obsession to Find the Lost Conqueror" (Basic Books, 2006).

Related: 2nd-century Alexander the Great statue with lion's-mane hairstyle unearthed in Turkey

It was still there around 300 years later, when the first Roman emperor, Augustus (reigned 27 B.C. to A.D. 14), visited Alexandria.

"He was desperate to view the corpse of Alexander," Cartledge said. "Apparently it was so visible that when he bent down, he knocked the corpse's nose off."

After that, the historical record makes no reference to the tomb for hundreds of years. "It disappears from mention around A.D. 500," Cartledge said.

Big gaps in the timeline

What happened to the tomb during the large gaps in the historical record has been a source of fevered speculation and even outright conspiracy theories ever since. These time spans also make it difficult to focus on specific theories about the tomb's current location.

Saunders thinks these interludes in the timeline are strange.

"It was only some 300 years after Alexander's death that we get an account [of the tomb]," while most records come from 600 years later, Saunders told Live Science in an email. "So, what happened in the intervening period? Why the silence?"

For hundreds of years, the tomb was a "very well-known pilgrimage center across the ancient Mediterranean world," he added, so at the time, the "location was not a secret."

So it's possible that no one bothered to describe the tomb's location because everyone knew where it was. There also may have been a rivalry between Christians flocking to Christ's tomb and pagans worshipping at Alexander's tomb. So once Christianity had the upper hand, Church fathers may have "made sure traces of the tomb were removed/damaged/destroyed so that Christ's tomb could then become the preeminent pilgrimage center of the ancient world," Saunders said.

Limneos-Papakosta thinks the site's location may have been lost during the Alexandrian Crusade in 1365, which ransacked the city, though no historical records suggest the tomb or actual body was destroyed, she said. However, one possibility is tied to the worship cult that arose around Alexander the Great. While it ostensibly died out thousands of years ago, there may have been people left in Alexandria who, during the crusade, still felt called to protect Alexander's remains from the incoming crusaders, who would have despised the worship of pagan gods.

"I believe that the priesthood around the cult of Alexander protected the tomb," Limneos-Papakosta said. "He was considered a god, and they tried to protect him when they realized that destruction was coming by Christians. I believe that they hid his body and sarcophagus."

Because Alexander's last known resting place was in Alexandria, most researchers agree that his tomb is likely hidden somewhere in the Egyptian city.

"Alexander is buried in Alexandria, and we are sure about that because in the first century A.D., we have information about visitors to the tomb, including Roman emperors like Julius Caesar," Limneos-Papakosta told Live Science. "So, we know that the tomb was there."

But exactly where in Alexandria is the question.

Shallalat Gardens

Limneos-Papakosta has one theory. During a 2009 excavation, she discovered a sculpture of Alexander, according to an article she wrote for Newsweek.

"It was our last day at the site, and we were ready to end the season," she said. "But then we found the sculpture. It was kind of a miracle."

Topographical maps and ancient sources suggest the area where the statue was found is close to where his tomb was once located, Limneos-Papakosta said, and her team has been excavating in the area ever since. Now, she thinks she's getting close, narrowing it down to the Shallalat Garden area of modern-day Alexandria, located near the Alexandria National Museum, which contains the remnants of the city's walls. Ancient sources suggest the royal quarters were in this location during antiquity, she said, yet they had never been excavated, she said.

The dig site is next to a former ancient intersection mentioned by second-century Greek writer Achilles Tatius, who describes the tomb's location.

"He wrote that it was a few hundred meters west of the intersection of Alexandria's two main broad streets," she said. Thanks to the current excavations, "we already knew of one of the broad streets [known as Canopic Street], but now we know of the second, Royal Street."

Under the sea

But others think Limneos-Papakosta is on the wrong track.

Cartledge, for his part, speculates that the body must have been housed inside a large mausoleum that was part of a royal complex. And he thinks the royal quarters were located in a different location than Papakosta-Limneos does.

"A mausoleum conveys the notion that it's a very solid structure that should be locatable," Cartledge said. "The one in Rome, Augustus' mausoleum, is completely visible. Why isn't the mausoleum of Alexander visible? The simple answer is that the Brucheum of Alexandria, where the royal quarters were, abutted the sea," Cartledge said. (The Brucheum was a particular quarter of Alexandria.)

But sea levels have risen several meters since the time of Alexander, historical maps suggest, so large swaths of the historical city, including the royal quarters, probably sit underwater.

In recent years, divers have explored the city's coastline and found pieces that could belong to ancient structures.

But even if a mausoleum is found, there's no guarantee a body would be inside. "Unless he was in a coffin that preserved his body, he will never be found," Cartledge said. "My guess is that his body could've been eaten by a shark."

Mazarita district

Saunders thinks the tomb is still on dry land, likely beneath the bustling streets of the modern Mazarita district, an area that was once the ancient city's hub and housed its palaces.

"Unless and until new building projects require demolition and clearance to a deep level, it is unlikely to be discovered," Saunders said. "For me, it is all but inconceivable it could be anywhere else."

Hawass doesn't have one site in mind, but he's optimistic that the tomb will be found one day.

"Can we say that the tomb can be discovered?" Hawass said. "Yes."

But it probably won't be found by archaeologists who are searching for it. The theater, catacombs and many of ancient Alexandria's monuments were found by accident, thanks to modern construction.

"The tomb will never be discovered by scholars — at all," Hawass said. "I really believe that by accident one day, the tomb of Alexander the Great will be discovered."

William Housty's grandparents taught him the sacred duty of preparing for the salmon's arrival each year. Before the first silver flashes appeared in the creek, his grandfather — following the wisdom passed down from his own elders — would clear woody debris, chase away seals, and maybe even fell a few trees to ensure a waterway was ready.

"They saw it as their responsibility to roll out a red carpet for the salmon because of their immense importance to us," said Dúqva̓ísḷa William Housty, a member of the Heiltsuk Nation of British Columbia's central coast.

This practice ensured that the salmon, the ecosystem and their community could thrive together, said Housty, who is director of the Heiltsuk Integrated Resource Management Department (HIRMD), which manages resources in their traditional territory.

Welcoming the salmon is just one example of the way the Heiltsuk's ancestral laws, or "Ǧvi̓ḷás — a set of principles centered on respect, responsibility, reciprocity and stewardship for all sentient beings — have shaped their interaction with their environment.

Now, the Heiltsuk are using traditional knowledge in concert with modern scientific approaches to monitor wildlife, count salmon, and maintain the health of waterways in their traditional territory. From the outset, the HIRMD stewards decided that Ǧvi̓ḷás would guide how they managed their resources, as well as influence how they would work with other government offices, industry or other outside parties.

This has led the Heiltsuk to braid relatively new techniques, like DNA analysis, with ancient ones, like the use of traditional fish weirs, so they can study — but not impact — the ecosystem. Their work has revealed shifting bear habitats and climate change impacts on salmon. Both have led to increased protections for creatures that are critical to the ecosystem.

"We're going back to the value system that our ancestors implemented for thousands of years," Housty told Live Science. "In our eyes, it is for the betterment of everything."

A symbiotic relationship

The Heiltsuk have lived in the diverse coastal rainforests, islands and marine areas of their traditional territory for more than 14,000 years. Over that time, they passed on ancestral knowledge of how to care for and enhance the natural resources they depended on.

In the mid-1800s, however, the British colonial government asserted control over Indigenous lands. In the following decades, deforestation, overfishing and pollution led to a marked decline in the richness of life.

"Take a look out there — it's beautiful," Housty said, pointing to the shimmering ocean water west of Bella Bella, the central community of the Heiltsuk Nation. "But when you go underwater, it's a different story — so many resources have been depleted to the extent that some of them have gone extinct."

For instance, commercial fishing has led to drastic declines in Pacific herring (Clupea pallasii), eulachon (Thaleichthys pacificus) and Northern abalone (Haliotis kamtschatkana). Some salmon that once thrived in the rivers and streams around Bella Bella have disappeared.

Government officials, forest companies and academia have their own agendas, said Q̓íx̌itasu Elroy White, an archaeologist who works part time for HIRMD. "It was often based on greed, commercial enterprise, and academic privilege and perspective," White said.

This contradicted the Heiltsuk way of living in harmony with the environment, in which they take only what they need to ensure a sustainable supply of resources for future generations.

Protecting the bears

For many decades after colonization, federal and provincial agencies controlled fishing quotas, logging operations and other resource management decisions that directly affected the Heiltsuk. However, that started to change in the 1990s, and a small team of Heiltsuk began doing field assessments on the health of the streams and salmon in the Koeye watershed, 34 miles (55 kilometers) southeast of Bella Bella. The team presented data to the Heiltsuk land use committee, which would use that information to craft conservation management plans. One key goal was to protect grizzly bear (Ursus arctos horribilis) habitat.

If "you protect grizzly bear habitat, you're protecting black bear habitat, wolf habitat, deer habitat and many other species," Housty said. "When you have lots of bears, it means you have a healthy ecosystem."

The Heiltsuk began monitoring bears directly in the Koeye watershed in the early 2000s.

The number of Heiltsuk researchers grew, and in 2010, the Heiltsuk formed HIRMD. That same year, they partnered with University of Victoria wildlife scientist Chris Darimont, who is also the science director at the Raincoast Conservation Foundation, and his graduate students. The academic team expanded the monitoring across a larger region of the Heiltsuk territory in a way that aligned with Heiltsuk values.

"A lot of the concepts were relational in that they were about how wildlife were like relatives to the Heiltsuk," Darimont said, "and ought to be treated accordingly."

Instead of catching grizzlies, sedating them and attaching tracking collars to them, which eventually fall off, the researchers took an entirely different approach: They created knee-high salmon-scented bear snares — barbed wire corrals around trees — and set 30 in the Koeye and more than 100 throughout the larger study area. Lured to the smell, the bears left hair samples, and the Heiltsuk used their DNA to track their movements. The noninvasive method didn't disrupt the bears' usual habits; the bait provided no rewards to the bears, so the grizzlies didn't become dependent on the snares for food.

Monitoring the bears, collecting scientific data and collaborating with academic scientists are proving critical to the Heiltsuk's involvement in management decisions, Housty said. Historically, when they didn't have the scientific resources, government organizations controlled the management of natural resources. "It is our partnerships and the science that has really given us the legs to stand on for joint management," Housty said.

The collaborations have helped the Heiltsuk identify bears on islands outside their traditional range and have pinpointed crucial corridors the bears use to move between feeding areas, according to new research that is under review for publication. Such findings have led to greater protection for bear habitats and will continue to do so, Darimont said.

The Heiltsuk stewardship principles often contrast with dominant government-led conservation approaches, which may suggest that hunting grizzly populations is acceptable when the numbers are sustainable. Seeing wildlife as a natural resource to be managed by humans is incorrect and unethical, Housty said. Even a single individual killed by trophy hunters is unacceptable, he said. "It violates our law regarding respect and reciprocity to the bears," Housty said.

In 2017, after gauging the broader public's views and listening to Indigenous perspectives, the British Columbia government ended trophy hunting of all grizzlies throughout the province.

Grizzlies feed on salmon during the spawning season, leaving the carcasses, skin, bones and leftover flesh to enrich forest soils and feed aquatic invertebrates, which, in turn, support juvenile salmon during their early life stages. Therefore, the ban on trophy hunting did more than benefit the bears; it strengthened the territory's salmon systems, Housty said.

Salmon stewardship

Each fall, Howard Humchitt and Lenard Stewart visit some of the many rivers on their territory, walking upstream to check salmon spawning habitats. They are Heiltsuk Coastal Guardian Watchmen, employed by HIRMD to take on many roles, from preparing for the salmon's return to monitoring the populations of chum salmon (Oncorhynchus keta) and pink salmon (Oncorhynchus gorbuscha).

Over the years, the watchmen have seen certain types of salmon diminish. Last year, the watchmen counted 7,000 salmon returning to a particular river system where they used to count tens of thousands. "When our fathers were kids, that same river probably had 100,000 salmon in it," Humchitt told Live Science.

The Heiltsuk historically used ancestral technologies such as fish weirs. These traps were made of wooden stakes driven into a river, which created a semipermeable barrier that directed salmon into a holding area as they swam upstream. The Heiltsuk also used stone fish traps — miniature stone walls that stretch across tidal inlets. Fish swam through high gaps in the walls and, when the tide receded, were trapped. However, the Canadian government outlawed such practices in the late 19th and early 20th centuries because officials believed the traps harmed fish stocks. But both technologies allowed the Heiltsuk to selectively harvest fish, so the best fish for breeding could go upstream and spawn. Having reliable fish counts can also prevent over-harvesting.

These systems have historically kept fish stocks stable, said William Atlas, a salmon watershed scientist at the Wild Salmon Center in Portland, Oregon. "Prior to the arrival of European colonists, there's probably about 7,000 or 8,000 years of successful stewardship of salmon harvesting," Atlas said.

In the mid-19th century, however, the colonial government of Canada took over fisheries management. Since then, overfishing, habitat destruction and nonlocal management have led to crashes in catches of salmon and herring in British Columbia. Several Chinook salmon (Oncorhynchus tshawytscha) populations in the province are at risk of extinction, and many coho salmon (Oncorhynchus kisutch) runs — migratory populations of a specific river system — are endangered. Atlas, along with Indigenous colleagues, has called for the revitalization of Indigenous fisheries by supporting the sustainable management practices of coastal First Nations.

In recent years, some coastal First Nations, including the Heiltsuk, have regained a stake in the management of their fisheries. Collecting data on their marine resources continues to be important to the Heiltsuk.

"Knowledge is power when it comes to ecosystems," Atlas said. "Having numerical values and abundance estimates of how many salmon are returning gives them authority when it comes to co-governance and decision-making."

To this end, the Heiltsuk have been monitoring and researching the current state of salmon systems in the territory to learn what it takes to sustain them. They also collaborate with scientists to collect data when necessary. It's logistically challenging to count sockeye salmon (Oncorhynchus nerka), as they spawn high up in river systems, typically above a lake. So the HIRMD researchers partnered with Atlas and his colleagues to monitor these populations, using an approach that braids Heiltsuk traditions in salmon stewardship with modern science.

Starting in 2013, the collaborators began using a fish weir made from locally harvested cedar logs to help them capture and tag around 500 sockeye salmon each year in the Koeye watershed. The researchers count the outgoing smolts — the young salmon swimming downstream on their way to the sea — and then install the weir, now made of aluminum, at the beginning of June and begin counting the incoming adults. During the salmon's brief capture, the team takes genetic samples from the fish so that when they are caught in the ocean, scientists can identify the salmon's population of origin.

Each year, the collaborators report sockeye numbers for five or more populations in the territory. The Heiltsuk can then share this information with Fisheries and Oceans Canada (DFO), the government agency that has historically set fishing quotas for commercial, recreational and Indigenous fishing in British Columbia.

"That creates a bit of a reversal in the power dynamic that we've historically seen, where the DFO is the broker of the truth," Atlas said.

Planning for the future

Looking out for future generations and acknowledging humans' role in ecosystems are core principles of the Heiltsuk's Ǧvi̓ḷás. Early in their conservation efforts, the Heiltsuk devised a 1,000-year natural resources management plan.

"Our goal is to live sustainably, so we can ensure an abundance of resources not only for my generation but for those to follow," Housty said. "We've been here since the beginning, and the long-term plan is to stay until the end — if there is ever an end."

Ǧvi̓ḷás also emphasize that humans are as responsible for caring for their territory as they are for their own homes. To uphold their stewardship practices, the Heiltsuk educate their children on their culture, fostering a connection with the natural world from an early age. The Qqs Projects Society, a Heiltsuk nonprofit in Bella Bella, supports youth, culture and the environment, offering programs to strengthen bonds with Heiltsuk lands and waters.

"We want our youth to feel connected to their territory because it's an intrinsic part of their identity," said Cúagilákv Jess Housty, executive director of the Qqs Projects Society. "And we want them to love it, because we know if they love it, they will protect it."

One goal is to demonstrate how humans can contribute positively to ecosystems. This requires embracing the Heiltsuk understanding that humans are not separate from their environment. Elders would tell Housty that the Heiltsuk didn't own their territory; it belonged to the animals and fish. That means that the Heiltsuk have a responsibility to take care of the creatures in their territory, William Housty said. "That's a far different mindset from viewing the land in terms of how much we can take," he said.

Reporting and travel for this story was supported by the Sitka Foundation and the Science Media Centre of Canada.

About 37,000 years ago, Neanderthals clustered in small groups in what is now southern Spain. Their lives may have been transformed by the eruption of the Phlegraean Fields in Italy a few thousand years earlier, when the caldera's massive explosion disrupted food chains across the Mediterranean region.

They may have gone about their daily life: Crafting stone tools, eating birds and mushrooms, engraving symbols on rocks, and creating jewelry out of feathers and shells.

They likely never realized they were among the last of their kind.

But the story of their extinction actually begins tens of thousands of years earlier, when the Neanderthals became isolated and dispersed, eventually ending nearly half a million years of successful existence in some of the most forbidding regions of Eurasia.

By 34,000 years ago, our closest relatives had effectively gone extinct. But because modern humans and Neanderthals overlapped in time and space for thousands of years, archaeologists have long wondered whether our species wiped out our closest relatives. This may have occurred directly, such as through violence and warfare, or indirectly, through disease or competition for resources.

Now, researchers are solving the mystery of how the Neanderthals died out — and what role our species played in their demise.

"I think the fact is, we do know what happened to Neanderthals, and it is complex," Shara Bailey, a biological anthropologist at New York University, told Live Science.

Decades of research has revealed a complex picture: A perfect storm of factors — including competition among Neanderthal groups, inbreeding and, yes, modern humans — helped erase our closest relatives from the planet.

The rise and demise of our closest human relatives

The modern story of Neanderthals began in 1856, when quarry workers found a strange-looking, not-quite-human skull in Germany's Neander Valley.

Archaeologists gave the skull a new species name: Homo neanderthalensis. And in the early decades after the discovery, researchers assumed the creatures were knuckle-dragging brutes. This depiction was based on a flawed reconstruction of a skeleton of an old Neanderthal man, whose spine was deformed by arthritis, found at La Chapelle-aux-Saints in France.

Now, more than 150 years of archaeological and genetic evidence makes it clear that these early human relatives were much more advanced than we originally thought. Neanderthals crafted sophisticated tools, may have made art, decorated their bodies, buried their dead, and had advanced communication abilities, albeit a more primitive kind of language than modern humans used. What's more, they survived for hundreds of thousands of years in the hostile climates of Northern Europe and Siberia.

Based on archaeological evidence from sites from Russia to the Iberian Peninsula, Neanderthals and modern humans likely overlapped for at least 2,600 years — and perhaps as long as 7,000 years — in Europe. That overlap occurred during a bleak period in Neanderthal history that ended with their downfall — raising the question of whether modern humans were responsible for killing them off.

But the story of Neanderthal life — and extinction — is one of regional variation, said Tom Higham, an archaeological scientist at the University of Vienna.

"In some areas, for example, we see that humans arrive to empty spaces in Europe where there aren't any Neanderthals anymore, seemingly," Higham told Live Science. "And in other places, we see that there's probably an overlap that happens … we know that people are interbreeding."

The first empirical proof of that interbreeding was found in 2010, when a Neanderthal genome was sequenced. Since then, genetic analysis has shown that Neanderthals and modern humans shared much more than a geographic area — we regularly exchanged DNA back and forth, meaning there is a bit of Neanderthal in every modern human population studied to date.

Related: 'More Neanderthal than human': How your health may depend on DNA from our long-lost ancestors

Already on the brink

When modern humans and Neanderthals met tens of thousands of years ago, the latter were probably already in trouble. Genetic studies suggest that Neanderthals had lower genetic diversity and smaller group sizes than modern humans, hinting at a potential reason for Neanderthals' demise.

"Genetically, one big clue that we get is this idea of heterozygosity," Omer Gokcumen, an evolutionary genomicist at the University at Buffalo, told Live Science. An individual receives two copies, or alleles, of a gene from each parent. Individuals are "heterozygous" for a given gene if they inherit a different allele from each parent. In Neanderthals' small communities, which contained fewer than 20 adults in each group, more inbreeding occurred. That meant fewer of them inherited different versions of a gene from each parent and, therefore, had low heterozygosity.

"Neanderthals may have suffered for that — what they call a mutational burden," Gokcumen said. Genetic research suggests Neanderthals had many problematic mutations that likely affected their survival. "Because of their small population size, they couldn't actually breed these bad alleles out, and their kids may actually be sickly at the end," Gokcumen said.

Any population of animals survives into the future through successful reproduction and rearing of offspring. Researchers estimating the mortality rates of Neanderthal infants have found that a decrease of even 1.5% in the survivorship of these children could result in population extinction within 2,000 years, April Nowell, a Paleolithic archaeologist at the University of Victoria in British Columbia, told Live Science.

"There doesn't have to be very much going on to really have a dramatic impact on the viability of your population," Nowell said.

So while Neanderthal populations began decreasing until they became small, isolated groups without the social support necessary to care for their increasingly sickly babies, modern human groups quickly expanded through Europe.

Humans and Neanderthals: A bad mix?

During two periods across Eurasia 135,000 and 100,000 years ago, Neanderthal populations nearly died out. But they bounced back, powering through these cold snaps and the resulting changes to their landscapes.

"Neanderthals survived all of these hardships," Bailey said. "It wasn't until they had the additional pressure of Homo sapiens that they eventually went extinct."

Given the overlap in time and space, researchers used to think modern humans played a direct role in Neanderthals' demise via warfare or novel diseases.

There is some evidence of violence on Neanderthal skeletons. A young adult male from St. Césaire, France, dating to 36,000 years ago suffered a fracture to the top of his head made with a sharp implement, and an older male found in Shanidar Cave, Iraq, dating to around 50,000 years ago had a partially healed stab wound on his left rib. But there's no way to say whether modern humans or other Neanderthals inflicted this violence. Unless archaeologists find a site where Neanderthals are clearly the victims of a massacre inflicted by modern humans, it will be impossible to conclude that modern-human violence was a major cause of Neanderthals' extinction.

Nor is there genetic evidence that modern humans' diseases killed off the Neanderthals, though we do share many immune-related genes. For instance, we inherited Neanderthal genes that make us susceptible to autoimmune disorders such as lupus and Crohn's disease as well as to severe COVID-19. Future genetic analyses may reveal the potential role of disease in Neanderthals' demise, Gokcumen said.

Winners and losers in the fight for resources

But warfare and pestilence aren't the only possible ways modern humans may have led to Neanderthals' vanishing. When two groups come together, competition can lead to tragic outcomes.

Neanderthal artifacts, such as pendants and etchings, show that Neanderthals were intelligent. But new research suggests there are significant differences between H. sapiens and Neanderthal brains: Modern humans have more neurons in brain regions key to higher-level thinking, and their neurons are more connected — meaning modern humans were likely more capable of thinking quickly. Combined with Neanderthals' greater difficulty in processing language, this could mean modern humans had an advantage in key tasks, Nowell said, such as hunting and foraging for food.

And while Neanderthals' extremely isolated groups may have had a biological disadvantage, they probably had a cultural one, too.

"Ideas can spread more easily when you have bigger populations and other people can build on them," Bailey said. But given Neanderthals' disparate populations, "their artistic or cultural kinds of innovations might not have progressed the way we see in much larger populations that have a lot of interactions with people," she said.

Although Neanderthals created tools that were very sophisticated for the time, we haven't found any unambiguous long-range weapons made by Neanderthals. Modern humans' ability to devise projectile weapons, by contrast, may have given us a survival advantage.

Related: Could Neanderthals talk?

But the full implications of these differences on Neanderthals' survival aren't yet known.

"We could also think about intergroup competition or competition between groups of Neanderthals," Nowell suggested, as a potential result of their dwindling numbers and modern humans' encroachment.

When looking at contemporary and historical hunter-gatherer groups, Nowell and co-author Melanie Chang, a paleoanthropologist at Portland State University, noticed that these groups often tightly regulate who has been able to use the land and its resources, and that being part of the "in group" can be a matter of survival. As Neanderthals began to disappear from most of Eurasia and retreat to southern Iberia, competition among Neanderthal groups would have risen.

"Maybe it's competition with other Neanderthals that prompts them to start differentiating themselves more," Nowell said.

That seems particularly persuasive, given that around 40,000 to 50,000 years ago, there was a cultural explosion among both modern human and Neanderthal groups. These cultural elements included a surge in personal adornment, such as painted shells, probably worn as pendants, that could have served as "group" symbols, Nowell said.

No cohesive, shared fate

Given the mounting evidence that Neanderthals and modern humans interacted regularly for thousands of years, many researchers are looking to an unusual place for the answer to what happened to Neanderthals: a theory first put forward by paleoanthropologist Fred Smith and colleagues 35 years ago.

"He came up with this suggestion that there was gene flow and a slow assimilation of Neanderthals into human populations," Higham said.

In essence, the two groups simply got used to hanging out with one another, and as more and more humans moved into Eurasia, eventually their larger population swamped the Neanderthals, whose lineage petered out. This idea is supported by a study that found H. sapiens simply absorbed the Neanderthals into our population. In that way, we may have made Neanderthals vanish as a distinct group — by making some of the remaining ones part of our family.

But this theory currently lacks "smoking gun" evidence of humans and Neanderthals living together for extended periods at the same site. They were sharing genes, but the archaeological evidence does not show Neanderthals and modern humans sharing either a home or the close social ties necessary to say we assimilated Neanderthals into our own population.

"Until they find a frozen Neanderthal and a modern human in a locked embrace, it's always going to be open to interpretation," Bailey said.

Even if we do find such a site, it is unlikely to change the nuanced, complicated picture of Neanderthals during their last moments.

"Some Neanderthal populations died out, some got massacred, some interacted and some only exchanged ideas," Sang-Hee Lee, a biological anthropologist at the University of California, Riverside, told Live Science. "The really exciting questions like 'Why did Neanderthals disappear?' 'Why did they go extinct?' no longer can have one overarching theory" she said. "Neanderthals as a whole did not have a cohesive, shared fate."

Milan's marble facades and narrow, stone-paved streets look elegant and timeless. But all of that stone emits heat and does nothing to absorb rain, and temperatures and flooding in the posh Italian city are only predicted to increase in the coming decades.

In Jakarta, black floodwaters already rush into homes every winter along the Indonesian city's many rivers. That water is filled with sewage and harbors disease, but many people can't afford to move. Soon, climate change will put more of Jakarta — and many other low-lying cities — below sea level.

And in arid San Diego, water is already treated like a precious commodity. As drought increases in the coming years, protecting this resource will become even more important.

Human-caused climate change is transforming weather patterns and shifting ecosystems around the globe. In some places, climate change means too much water; in others, it causes drought. Global action is needed to curb fossil fuel use, slow the rise in temperatures and prevent the worst impacts of human-driven climate change. But significant warming is already baked in.

Cities will have to respond, and some are already taking bold steps. Milan is planting millions of trees. Indonesia is moving its capital city. And San Diego is recycling wastewater back into city taps — one of the first major cities to do so.

Each of these three cities offers a different roadmap for climate adaptation that has lessons for other places around the world. And while no single approach will be a silver bullet, each offers a hopeful vision of how we can learn to live and thrive on a warming planet.

In Milan, for example, the city is working to plant trees throughout the city, rather than just focusing on the wealthiest areas.

"I think it stands out as a successful role model that other cities can learn from," Matilda van den Bosch, a researcher with the European Forest Institute, told Live Science.

Milan: A forest in the city

Like many cities, Milan is a "heat island": Temperatures there are 7 to 14 degrees Fahrenheit (4 to 8 degrees Celsius) hotter than in surrounding rural areas. This is because buildings, roads and other infrastructure absorb and reemit heat from the sun better than forests and bodies of water do. And climate modeling predicts things will only get worse, with temperatures in the city rising by up to 4.1 F ( 2.3 C) by 2050.

To address this threat, the city has launched a public-private partnership called ForestaMi — or Forest for Milan — that aims to plant 3 million trees and bushes by 2030. As of 2024, it had planted more than 610,000 trees and bushes. The tree-planting initiative is part of a larger climate plan that Milan hopes will help it keep local warming below 3.6 F (2 C) by 2050.

In the urban planning world, planting trees is a popular climate mitigation strategy. Trees and other vegetation bring down temperatures by offering shade, absorbing and diffusing heat better than cobblestones and pavement, and releasing moisture into the air. Increasing tree canopies over European cities could save thousands of lives by blunting the impact of urban heat waves, according to a 2023 study in the journal The Lancet.

But planting trees may have other benefits, too. Replacing pavement with soil can help cities absorb more rainwater and reduce flooding. That will prove essential in Milan, which climate modeling predicts will face more torrential rain in the coming decades.

But tree planting has limitations. In July 2023, a sudden hailstorm hit Milan, downing 5,000 trees in just 15 minutes, city council member Elena Grandi told Live Science in an email. While storms like this are rare in Milan, Grandi noted, the city will face more river flooding and drought in the future, meaning it will need a mix of trees that can withstand such conditions.

"We have learned that it is necessary to plan urban green spaces in a different way, planting varieties more resistant to storms or to extreme temperature and water scarcity," Grandi said.

Jakarta: Mass relocation

Jakarta, a megacity roughly eight times larger than Milan, faces both too little and too much water. Sea level rise is already a crisis in low-lying Jakarta, a swampy city that is crisscrossed by 13 rivers.

Jakarta was built during the Dutch colonial era around a series of canals that never quite contained those rivers. Since 1990, Jakarta's population has more than doubled, further straining the city's infrastructure. Because Jakarta cannot provide enough piped water to its residents, owners of both high-rise condos and informal shacks dig illegal wells to pump groundwater, said Deden Rukmana, a professor and expert on Indonesia's urban planning at the University of Texas at Arlington, told Live Science.

This overpumping helped make the city one of the fastest-sinking areas in the world; some places are dipping 4 inches (10 centimeters) per year. Coastal flooding forces people out of their homes, and at current rates, 95% of the city's coastal district is estimated to be underwater by 2050.

And the risks are not confined to the coast. The combined effects of sea level rise and groundwater depletion will put the entire city below sea level by 2100, modeling predicts. Rising seas also mean that salt water will contaminate the city's freshwater supply.

To buy some time, Jakarta's city planners are already building and reinforcing dikes and estuaries closer to the shore. As a next step, they envision building a cluster of 17 islands shaped like a giant bird. Together, those islands would create an 80-foot-tall (24 meters), 25-mile-wide (40 kilometers) seawall and an artificial lagoon that planners hope will help buffer the city from tidal flooding.

But researchers warn that even a huge seawall won't prevent flooding if overpumping continues to drive subsidence inland. Efforts to relocate water-guzzling industries and to promote development in areas that are less prone to flooding are helping on that front, Rukmana said.

To stay above sea level, Indonesia's primary city will need to crack down on illegal wells and create alternate water sources, which could take years, Rukmana said. The city could also work to refill aquifers, as Tokyo has done.

But there is another step that could relieve pressure on Jakarta's groundwater: mass relocation. In 2019, Indonesia announced plans to move its capital from Jakarta to a new city called Nusantara, on the island of Borneo. The first civil servants are expected to move there in September 2024.

Like similar relocations in Brazil and Nigeria, Rukmana said, the project aims to relocate a colonial-era capital to a more central location within the country.

"Without any intervention, people will still move to Jakarta," Rukmana said. Nusantara won't change that right away, but it could have an impact over time. When Nusantara is built, the goal is that 10,000 civil servants will have their jobs relocated, he said.

Of course, relocating doesn't always go as planned. In 1999, Malaysia moved its prime minister's office from Kuala Lumpur to nearby Putrajaya, also due to water issues. Over the past two decades, its population has increased to 100,000 — but that number is less than one-fifth of the total population envisioned. Rukmana said building a new city from scratch is a big risk but one that could pay off for Indonesia if it provides a new source of development for the country.

So far, Nusantara is mostly just cleared land. But planners envision a smart city built around public transportation, walkable neighborhoods and electrification, with robust digital tools for management — a far cry from Jakarta's congested traffic, air pollution and overcrowding.

Indonesia recently partnered with the United Nations to get input from people living in Argo Mulyo, an existing village that will be incorporated into Nusantara's footprint.

Nusantara also has tree-planting goals, which are arguably even more ambitious than Milan's. The new city will be located on land that was previously used for industrial agriculture; it aims to reforest 204,000 acres (83,000 hectares) of rainforest. At the city's current planting densities, that would be the equivalent of tens of millions of trees.

So far, however, experts say those efforts are falling short, Mongabay reported. Non-native trees, such as Eucalyptus, are being planted instead of local rainforest species, and a phased approach may be needed to account for the area's damaged soil. For example, "pioneer" tree species may need to be planted first, followed by typical rainforest species, to replicate natural forest processes.

Nusantara city leaders say they are working on a master plan to address these challenges. In the long run, however, Jakarta still needs to address its own problems, Rukmana said.

San Diego: No-waste wastewater

While low-lying Jakarta struggles with flooding, arid San Diego faces drought. The California city, which gets less than 12 inches (30 cm) of rainfall a year, is predicted to experience hotter temperatures and less-predictable rain as climate change worsens.

But San Diego is in a better place than many cities because its climate adaptation actually began decades ago. When California faced a drought in 1990, San Diego's water supplier, the Metropolitan Water District of Southern California, temporarily cut the city's water supply by half.

"The political leaders in San Diego said, 'We just can't have this; we need to create our own water independence,'" Jeff Stephenson, now the director of water resources for the San Diego County Water Authority, told Live Science.

At that point, they began rolling out a low-flow-toilet initiative to reduce household water consumption that inspired similar efforts around the country. But San Diego was just getting started.

Over the past 30 years, the city has pulled just about every lever available to reduce demand. Low-flow toilets were followed by low-water landscaping and a water conservation partnership with inland agricultural areas.

Since it started conservation efforts, San Diego County has successfully halved its per capita water use and reduced its reliance on water from Los Angeles by even more. But water conservation on its own isn't enough. "You can't conserve your way completely out of a drought," Stephenson said. "You're still going to need water."

So, the water authority built and raised dams for more water storage and lined canals to prevent water from seeping away en route from the Colorado River. It built pumps to move water north within San Diego County, partly in case of droughts.

And San Diego also moved to beef up its local water supply. Because the city's plentiful ocean water makes up for its lack of groundwater, San Diego invested in the nation's largest desalination plant, which currently supplies around 10% of the region's water.

After three decades of work, plus a couple of wet winters, San Diego now has water to spare. It's even planning to lease some of its Colorado River supply to nearby cities.

And facing the likelihood of more climate-driven drought, the city is still aiming to do more. For that, San Diego is again looking to toilets for inspiration. Over the next decade, the city of San Diego and two suburbs plan to recycle wastewater back into city taps. It goes beyond gray water — not-quite-drinkable water that's used for landscaping, or indirect reuse, which mixes wastewater into other reservoirs first. Instead, San Diego wants to send purified wastewater directly back into the city's potable water supply. This is a new approach in California.

That level of purification may sound daunting, but when planners ran the numbers, it was cheaper than other waste upgrades, Stephenson said. San Diego County expects 18% of its water to come from recycled water by 2045.

All of these changes took political will. They also required substantial resources: San Diego benefits from a stable population, plenty of funding and efficient government systems, which many cities worldwide lack.

And for San Diego — as well as Milan, Jakarta and Nusantara — one moonshot project will not be a silver bullet for avoiding climate impacts.

Stephenson encouraged cities to use a variety of approaches — and to stick it out. Change happens one step at a time. "It's a slow process," he said, but as in San Diego, the results eventually come.

Editor's Note: This story was updated on Tuesday, Sept. 24 at 9:30 a.m. EDT to update Deden Rukmana's affiliation after a recent move and to correct the name of the water authority that supplies San Diego's water; it is the Metropolitan Water District of Southern California, not Los Angeles. The article was also updated to clarify that San Diego's water recycling project goes beyond indirect reuse.

For about seven minutes in 2003, scientists reversed extinction.

The resurrected lineage was the Pyrenean ibex (Capra pyrenaica pyrenaica), and the last known member of the subspecies, a female named Celia, had died three years earlier.

Scientists had collected DNA from Celia's ear before her death and injected her genetic material into a domesticated goat egg cell with its nucleus removed. The resulting clone — the first and only extinct creature to have been revived at the time — died soon after birth due to a lung defect.

Although that effort failed to produce a healthy animal, "de-extinction" science has advanced dramatically in the past two decades. Technology is no longer a significant hurdle to reviving recently extinct species, and in many cases, we have enough DNA to piece together functional genomes for cloning. The question isn't so much whether we can resurrect lost species but if we should.

Some companies aren't waiting to answer that question. For instance, Colossal Biosciences, a Texas-based biotechnology and genetic engineering company, plans to bring back three iconic extinct species: the dodo (Raphus cucullatus), the Tasmanian tiger (Thylacinus cynocephalus; also known as the thylacine) and the woolly mammoth (Mammuthus primigenius). The ultimate goal for these de-extinction efforts, according to Colossal's website, is to "enrich biodiversity, replenish vital ecological roles and bolster ecosystem resilience."

But a catastrophic outcome cannot be ruled out, other experts say.

"We have this hubris as humans that we can control our technology," Oswald Schmitz, a professor of population and community ecology at Yale University, told Live Science. "I'm not so convinced."

Even in the best-case scenario, conservationists are skeptical that bringing back creatures that died out centuries or millennia ago will offer as much benefit as preserving the ones that are still hanging on.

"What's gone is gone," Schmitz said.

Closer than people think

Recent breakthroughs have brought scientists "closer than people think" to reviving long-extinct species, Ben Lamm, co-founder and CEO of Colossal Biosciences, told Live Science. The company aims to produce its first mammoth look-alike calves by 2028, and "it is highly likely that one could see another species before then," Lamm said, referring to Colossal's other de-extinction projects.

Woolly mammoths lived in the Arctic between 300,000 and 10,000 years ago. To produce the calves, Colossal scientists will first identify the genes encoding the woolly mammoth's most emblematic physical traits, such as shaggy hair, curved tusks, fat deposits and a dome-shaped cranium. They will then insert these genes into the genome of closely related, and therefore genetically similar, Asian elephants (Elephas maximus).

"De-extinction" can have different meanings — and our capability for de-extinction depends on how we define it, Love Dalén, a paleogeneticist and professor of evolutionary genomics at Stockholm University, told Live Science in an email. De-extinction in the sense of "creating a form of hybrid species that looks like the extinct one by replacing a limited number of key genes" is possible if there is available DNA from the extinct species, said Dalén, who sits on Colossal's advisory board. (There is no available DNA from dinosaurs, so creating a Jurassic Park is currently more of a pipe dream.)

Creating an animal that is genetically identical to the extinct species is trickier and depends on the species. For species that went extinct very recently and for which there are high-quality DNA samples, like Celia, "this could be possible," Dalén said. That's not the case for mammoths, however.

Dalén and his colleagues have come close to sequencing complete woolly mammoth genomes, but a few regions of DNA, such as some strings of repeat code, still prove challenging.

So far, Colossal Biosciences has acquired more than 60 partial woolly mammoth genomes, which it will use to edit elephant genomes in the lab, Lamm said. Once they have settled on a DNA sequence, Colossal representatives say they will implant a hybrid elephant-mammoth embryo into an Asian or African elephant (Loxodonta) surrogate.

The company also plans to resurrect the dodo and thylacine. "On the dodo project, we have one nearly complete genome," Lamm said, while on the thylacine, which only went extinct last century, "we've got much better genetic data right out of the gate." For the dodo, Colossal plans to implant the genetic material into domesticated chickens (Gallus domesticus) and, for the thylacine, into mouse-like marsupials called fat-tailed dunnarts (Sminthopsis crassicaudata).

Filling a gap in the ecosystem

The goal of de-extinction is to fill ecological niches that have stood empty since the original species disappeared, said Ronald Goderie, an ecologist and the director of the Taurus Foundation, a branch of Rewilding Europe that aims to create a substitute for extinct aurochs (Bos primigenius). Aurochs, the wild ancestors of domestic cattle (Bos taurus), once roamed across North Africa, Asia and nearly all of Europe. They likely played a vital role in maintaining animal and plant biodiversity through grazing and trampling the soil, according to the Taurus Foundation.

Goderie and his colleagues are bringing aurochs back through back-breeding, which doesn't involve genetic engineering. Humans hunted aurochs to extinction in 1627, but their DNA lives on in ancient cattle breeds living in Southern Europe, Goderie told Live Science. By selecting and breeding together cattle with physical attributes, behavioral traits and genes similar to those of aurochs, ecologists are inching toward reviving the lost species.

"We're very close," Goderie said. "We see very large, steep progress with each generation."

Woolly mammoths, which thrived during the Pleistocene epoch, maintained Arctic grasslands by trampling the snow, suppressing tree and shrub growth, and dispersing nutrients across huge distances via their dung. Without mammoths and other now-extinct megaherbivores, the so-called "mammoth steppe" has given way to a waterlogged landscape of mossy tundra, shrubs and forest. Research suggests this new landscape stores less carbon than grasslands did and is more susceptible to permafrost thaw. Some scientists have proposed that reintroducing megaherbivores to the Arctic could help restore the ecosystem, boost carbon storage and mitigate climate change.

The dodo was one of the largest terrestrial animals in its ecosystem in Mauritius. Evidence suggests dodos were seed dispersers and thereby influenced vegetation growth — although some experts question the extent to which the birds shaped their ecosystem. Colossal's plans to reintroduce dodos include ridding Mauritius of invasive species so that the birds can thrive, which the company argues has a beneficial "halo effect" for other species, Lamm said. The company is working with the Mauritian Wildlife Foundation, a non-governmental organization, and intends to partner with the Mauritian government, according to its website.

As for the thylacine, it was Tasmania's only marsupial apex predator, according to the University of Melbourne's TIGRR Lab, which is also working on de-extinction. As such, the species was important for predator-prey relationships that stabilized the ecosystem.

Scaling up

But de-extinct species can fulfill their ecological roles only if they survive in the wild and their populations are large enough, Goderie said. In the case of mammoths, that's a behemoth task.

Researchers recently estimated that Alaska's vast North Slope could support 48,000 woolly mammoths.

But to fulfill their ecological role — which may include increasing the albedo effect, where snow cools Earth by reflecting light back into space — mammoths would likely need to be reintroduced across the Arctic, said Vincent Lynch, an evolutionary biologist and associate professor at the University at Buffalo.

"You can't just put them in Alaska and fix the permafrost problem; you have to put them everywhere," Lynch told Live Science.

There would need to be several hundred thousand mammoths to have a significant impact on the climate, and that could put living endangered species in peril, Lynch said.

Both Asian and African elephants, which Colossal plans to use as surrogates to grow mammoth calves, are endangered, and every elephant gestating a "mammoth" calf can't grow babies of its own. "That's going to reduce population size," Lynch said.

The alternative would be to implant embryos into artificial wombs, but those have yet to be fully developed. "There would be like hundreds or thousands of artificial uteruses in a warehouse somewhere, churning out genetically engineered mammoths," Lynch said. "This is like '1984' scary stuff."

Unintended consequences

In addition to facing technological hurdles, de-extinction could have serious repercussions. For one, de-extinct animals may be sickly, given that the pool of available DNA for each species is relatively small.

Sustaining populations requires a sufficient number of creatures that are genetically different enough to buffer against diseases and harmful mutations.

It's also worth considering who would be liable if large-scale mammoth reintroductions went wrong. "The ecosystem has been adapting to the absence of mammoths since mammoths started going extinct," Lynch said. "What if there's an unintended consequence and something bad happens?"

Other experts echoed these concerns. "To get some impact, you need to have a lot of animals," Sophie Monsarrat, an ecologist and the rewilding manager at Rewilding Europe, told Live Science.

Reintroductions can lead to clashes between humans and wildlife. "In Africa, if you look at elephant conservation and reintroductions, then there are conflicts," Monsarrat said. In Kenya alone, human-elephant conflicts killed 200 people between 2010 and 2017, according to WWF. If scientists were to succeed in creating a viable mammoth population, they would have to roll out continent-wide education programs "teaching people how to react in front of a mammoth," Monsarrat said.

The makeup of the reintroduced population matters too, Schmitz said. Brown bears (Ursus arctos) that were brought from Slovenia to the Italian Alps in the 2000s are thriving, but it turns out that the source population was particularly aggressive, Schmitz said. Ecologists "didn't necessarily do their homework," he said. "They reintroduced them, and now they're attacking livestock and people."

There's also no guarantee that animals will stay where we release them, Schmitz said. In 1995, gray wolves (Canis lupus) were reintroduced to Yellowstone National Park but have roamed far beyond the park's boundaries. "If you unleash something in nature, these animals are going to find places that are the most suitable for them and may not be the places where you think they will stay," Schmitz said.

And although megaherbivores helped with carbon storage during the last ice age, "restoring megafauna may just as well have a negative impact on climate change," Dalén said. For instance, mammoths could theoretically contribute to global warming by eroding the permafrost during the warm season and by emitting methane, he said. They could also reduce the amount of carbon stored in woody plants, as elephants do when they eat plants in African savannas.

"A pet project of billionaires"

While proponents think de-extinction is a valuable conservation tool, critics say money would be better spent on existing initiatives. Colossal Biosciences has raised at least $225 million to fund its de-extinction programs, with some of that money also going toward conservation and genetic research across the globe, Lamm said.

Instead of using that money to bring back three extinct species whose ecological impact is unknown, the funds could be put toward saving roughly 100 species that are currently facing an uncertain future, Schmitz said.

In some cases, de-extinction targets species with a living equivalent fulfilling similar ecological functions. For example, one reason the thylacine went extinct was that people perceived it as a predator of livestock, Schmitz said. "But we have a similar animal right now that's facing similar persecution, and that's the dingo [Canis lupus dingo]," he said.

Lynch agreed that the best way to support conservation is to spend money on existing, proven initiatives. It's probably impossible to introduce an ecologically significant number of de-extinct species, which also raises questions about the purpose of de-extinction companies like Colossal, he said.

Adam Searle, a research fellow in cultural, historical and environmental geography at the University of Nottingham in the U.K., said de-extinction is unlikely to contribute to solving the ecological crisis. "It's literally a pet project of billionaires," he said.

In response to Searle's comments, Lamm told Live Science in an email that "for every de-extinction target, the technology and techniques developed will benefit closely related species." He further argued that "technological innovations made and funded by private sector companies are vital to this fight."

Regardless of Colossal's ecological impact, the company could still see a big return on investment, Lynch said.

"If they're successful — and I have no doubt that they will be — they're going to make a ton of money," he said.

For the rest of us and for nature, however, de-extinction "could do a lot more harm than good," Schmitz said.

In 2021, a man in his 50s was transferred to the intensive care unit at Emory University Hospital in Atlanta. He was in septic shock due to a bacterial infection that had spread to his bloodstream. The culprit? Klebsiella pneumoniae, which was resistant to most antibiotics.

Finally, hope emerged: A two-week course of an antibiotic called cefiderocol seemed to clear the infection. But just 10 days later, the man was rushed back to the ICU, where doctors discovered a pus-filled mass covering his liver. The same bacteria had returned with a vengeance.

The doctors sent the patient's blood sample to David Weiss, a microbiologist at Emory Antibiotic Resistance Center. Weiss discovered that the bacteria were now highly resistant to cefiderocol.

He quickly notified the doctors. But even after switching antibiotics, the man died.

The man's case highlights an elusive strategy that bacteria use in their arms race against antibiotics — a hidden type of resistance that can be switched on almost instantly but leaves no genetic trace. And it can be almost impossible to detect using standard lab tests. Researchers are increasingly recognizing that this bacterial strategy, called "heteroresistance," may play a significant role in antibiotic failures. Recognizing this bacterial strategy, experts say, is the first step in fighting it.

"While this is a worrisome phenomenon that we're studying now, in the big picture and in the future, I think we'll be able to use this knowledge to greatly help patients," Weiss told Live Science.

RELATED: Superbugs are on the rise. How can we prevent antibiotics from becoming obsolete?