NASA named the international crew of four that will fly on its next Artemis mission as early as next year. It's a key test flight ahead of a human landing mission to the lunar surface that the agency calls "one of the most highly complex missions NASA has undertaken."

The Artemis III mission, while vital to future moon missions, will remain closer to home in low-Earth orbit. There, it will demonstrate the Orion spacecraft's rendezvous and docking capabilities with two commercially designed and built lunar landers.

NASA astronaut Randy Bresnik will command the mission. A member of the U.S. Marine Corps, Bresnik has flown to space twice and logged 149 days off the planet. European Space Agency astronaut Luca Parmitano, from Italy, will serve as the mission's pilot. Parmitano has also been to space twice, including one dangerous spacewalk that was cut short when he nearly drowned as his helmet filled up with water.

NASA's Frank Rubio and Andre Douglas will serve as mission specialists. Rubio served in the U.S. Army and is a board-certified family physician and flight surgeon. He has flown to space once, to the International Space Station on a Russian Soyuz capsule. After engineers discovered the spacecraft was damaged, Rubio's stay was extended while awaiting a new ride, and he set the record for longest spaceflight by an American at 371 days.

This will be the first spaceflight for Douglas. The Coast Guard reserve officer was selected as a NASA astronaut in 2021. He is a systems engineer with a doctorate from George Washington University. Douglas served as backup crew to Artemis II.

"This mission is going to be fantastic," he said at the crew's announcement Tuesday from NASA's Johnson Space Center in Houston, Texas. "What an excellent crew. Got Artemis! Go NASA!"

Critical flight test

Artemis III is a crucial mission as NASA tests key pieces of hardware that will return humans to the lunar surface as early as 2028 — and for the first time since 1972.

To accomplish that, the agency is relying on lunar landing vehicles that are designed and developed by private companies SpaceX and Blue Origin.

Artemis III will test the ability of the Orion crew capsule to rendezvous and dock with those landers not near the moon, but instead in low-Earth orbit.

"This gives our teams key information on systems the lunar lander crew will depend on in an environment close to home versus four plus days away around the moon," said NASA's Jeremy Parsons.

The mission begins with the uncrewed launch of Blue Origin's Blue Moon lander, which can stay in orbit for 90 days. That gives NASA time to launch the crew in the Orion space capsule atop NASA's SLS rocket from the Kennedy Space Center in Florida.

After launching, the crew will pilot the Orion spacecraft to catch up with the lander and dock to it. There it will remain for about two days, and the crew will open the hatch and test key elements of the lander — like its life support and control systems — before undocking.

SpaceX will then launch its Starship into low-Earth orbit. Orion and the Artemis III crew will dock with that lander, spending about a day attached to that spacecraft.

Then, Orion and its crew will come home, splashing down in the Pacific Ocean off California.

The path ahead

NASA administrator Jared Isaacman wants to launch this mission by the end of next year. But an accident last month at Blue Origin's launch facility at Cape Canaveral Space Force Station in Florida brings some uncertainty to that timeline.

During a routine, prelaunch test, Blue Origin's New Glenn rocket exploded at its launch site, causing significant damage to the facility. Blue Origin's lunar lander for the Artemis III mission is scheduled to launch on New Glenn.

"While we recognize there are questions about how Blue Origin's recent anomaly impacts our plans, setbacks are a learning opportunity," said Blue Origin's John Couluris. "We are confident that New Glenn will be ready for Artemis III."

SpaceX is also working to develop its lander, completing a 12th test flight of the Starship system. The company has yet to launch Starship into orbit.

As the two companies prepare their landers for the Artemis III test, the crew is working with the Artemis II astronauts to prepare. "The most important Artemis mission will always be the next Artemis mission," Bresnik said.

At the crew announcement, the Artemis II astronauts handed over a baton — one that was carried during their mission around the moon and back — as a symbol of that step-by-step development that will help NASA achieve its ambitious goal of returning humans to the moon.

"We've been carrying these batons around for way too long," said Artemis II commander Reid Wiseman. "So with that, the Artemis II crew hands you the baton. You got the controls."

Copyright 2026 NPR

Apple took a leap into the AI era on Monday, announcing a long-awaited update to its digital assistant Siri and changes in its operating systems that attempt to more deeply integrate artificial intelligence.

The company's share price fell close to 2% after the news, though. While analysts say the changes have potential, whether or not they are a hit with consumers will have to wait to be seen once they are made available to the public later this year.

The Siri overhaul comes after repeated delays that raised questions about Apple's commitment to AI, as chatbots and agents have taken center stage in the tech world amid a tsunami of AI investments by other companies.

"Today we're taking a big step forward," Craig Federighi, Apple's senior vice president for software design, said at the company's Worldwide Developers Conference, or WWDC, held in Cupertino, Calif. He defended Apple's approach, stressing the company's focus on utility and protecting user privacy.

"Some appear to be racing forward, seemingly pursuing AI for the sake of AI without clear regard for the people — all of us — that it's ultimately meant to serve," he said. "We believe that truly helpful AI must be centered around you and your needs."

As chatbots like OpenAI's ChatGPT and Anthropic's Claude have wowed consumers with their ability to answer complex questions and perform digital tasks, Siri has lagged far behind. On Monday, Apple put on display its new version, with a string of videotaped demonstrations in which Apple employees showed off Siri's new features.

The company said the new Siri, dubbed Siri AI, will be accessible through a standalone app, and also through the search function on the home screen of Apple devices and in various other apps, like Photos. It will be able to access cloud computing networks and the internet, but be informed by a user's personal experiences and information on the user's Apple devices, such as their email and text message history.

Apple executives showed how Siri AI could find information online, give recommendations for things like menus, dig into texts or emails to pull up addresses or other information, and move photos into albums.

A 'prove-it moment' for Apple

Ben Bajarin, CEO of the tech research company Creative Strategies, said there was a low bar for making improvements, given Siri's current limited functionality. He said Apple is in a good position to bring AI to a broad customer base through its popular products.

"It looks like it's a pretty big upgrade," he said of the new Siri and Apple's deeper integration of AI into its systems. "It will work very cleanly into a whole slew of things that consumers might already do."

But, he added, "I think now we'll just have to see how it actually works."

Francisco Jeronimo, an analyst with the consultancy IDC, said Apple appears to want AI to "disappear into the operating system," rather than make chatting with an AI the user's focus.

"If Apple makes AI feel natural, private and useful for mainstream users, it will not just strengthen its ecosystem. It could redefine what consumers expect from every device they use," he said.

Presenters at the conference said Siri AI would be available to U.S. customers later this year in English, with other languages coming soon. It will not be immediately available in the European Union or China, two big markets for Apple, due to international regulations.

Apple is turning to one of its biggest phone hardware rivals for help catching up in AI. In January, Google and Apple announced a multi-year collaboration under which Google's Gemini AI model would be the basis for Apple's AI systems.

Daniel Newman, CEO of the Futurum Group, a tech research and advisory firm, said Apple now faces a "prove-it moment."

"My first reaction — and I think the reason the stock sort of fell — was it's ticking a box, but still uninspiring," Newman said.

"Given the fits and starts of Apple's AI rollout over the last few years, I don't know that they've given us enough reason to believe they can be trusted this time. The proof is going to have to be in the delivery, in the execution," he added.

Newman said the promises Apple made on Monday have potential, and there is clearly a big market for an improved Siri. Investors may also like the fact that Apple is "just paying rent to Google" for Gemini, he said, rather than shoveling money into self-made AI development, like many other tech companies are doing.

Beyond AI, Apple announced new parental controls on app and content access for kids, including the ability to limit which websites they can browse and apps they can download, who can communicate with them, and how much time they can spend on devices.

These updates come at a time when a range of AI and social media companies are facing a slew of lawsuits claiming harms to minors, including mental health troubles and exposure to violent or sexually graphic conversations with chatbots.

Tim Cook's last WWDC as Apple's CEO

Outgoing Apple CEO Tim Cook opened and closed the main session with brief remarks at the start of what will likely be his last WWDC, but did not play a role in unveiling the new products.

In April, Cook announced that in September he will hand the reins of a company that he helped make one of the world's most valuable to John Ternus, a mechanical engineer by training who currently oversees the development of Apple hardware, like Mac computers and iPhones.

In an April letter announcing the leadership change, Cook lauded Ternus as someone with "the mind of an engineer, the soul of an innovator, and the heart to lead with integrity and honor."

Despite the perception that Apple has had missteps on AI, the company has thrived under Cook, with its share price soaring some 2,000% on a split-adjusted basis during his 15-year tenure as CEO. Cook expanded the array of revenue-generating services that Apple offers, with products like Apple Pay, Apple Music and Apple News+. He also launched a line of custom microchips to power Apple's products.

Still, Cook has taken flak for binding Apple's fate to China as its manufacturing hub, a move that created supply chain efficiencies but has become a political risk that the company is now trying to address through diversification.

Critics have also said Cook lacks his predecessor Steve Jobs' ability to drive wow-factor product innovation, instead delivering a string of incremental device updates over the years.

Closing out the morning keynote on Monday, Cook said "the best is still ahead" for Apple, which he said strives to create the best products to deliver enriching experiences.

"It's been the honor of a lifetime to help advance that mission with teams whose creativity, care and conviction continue to make a lasting difference in peoples' lives," he said.

Note: Apple and Google are financial supporters of NPR.

Copyright 2026 NPR

Gray whales used to be a rare sight in San Francisco Bay. Now, their spouts are appearing off Alcatraz Island in one of the busiest waterways in the country.

The whales are making a pit stop on their long migrations from Mexico to Alaska, detouring under the Golden Gate Bridge for a snack as climate change is shrinking their normal food supply in Arctic waters.

But as gray whales try to adapt to one human-caused impact by feeding in San Francisco Bay, it's putting them squarely in the path of another hazard: ships.

Of 16 gray whales seen in San Francisco Bay this year, seven have died. Researchers have found evidence that several were killed by ship strikes.

With some whales now hanging out in the bay for weeks, a coalition of marine scientists and local officials is trying out a new system to prevent collisions.

Researchers installed a thermal camera on an island in the bay that can spot heat from the whales' exhalations. Potential whale sightings are screened by artificial intelligence and then confirmed by human screeners. The U.S. Coast Guard can then use that information to alert vessels and ships.

"We want the word to get out," says Gary Reed, director of Vessel Traffic Service San Francisco for the U.S. Coast Guard. "We want people to know there are whales in a particular location so they don't encounter them."

Gray whales in the North Pacific are declining. The current population is around 13,000, half of what it was a decade ago. Last year, 22 gray whales died in the larger San Francisco Bay Area, the highest number in 25 years. The same is happening in other areas along the West Coast.

"We're looking at a moment for gray whales where every whale that comes in and goes out of the bay matters for population," says Douglas McCauley, director of the Benioff Ocean Science Laboratory at the University of California, Santa Barbara. "So even though this is just one piece of the problem, it's a piece that we want to solve, can solve."

Climate change ripple effects

On a beach on Angel Island, the giant vertebrae of three whales lie in a row on the sand. They're the remnants of whale necropsies — animal autopsies done by two Bay Area research institutions, the Marine Mammal Center and the California Academy of Sciences. When a dead whale is spotted, the researchers rapidly try to assess the cause of death.

Broken bones and bruised tissue are often a sign of a ship strike.

That's visible in the middle skeleton, from a female that came into the bay this year, says Kathi George, director of Cetacean Conservation Biology at the Marine Mammal Center.

"She died from injuries due to blunt force trauma from vessel strike," George says.

In recent years, an alarming number of whales have also been washing up malnourished, both in California and in the Pacific Northwest.

Gray whales undertake one of the longest migrations of any mammal, traveling around 12,000 miles round trip every year. They spend the summers feeding in the cold waters of the Arctic, where prey is abundant. Then they swim to Baja California, Mexico, for the winter, where they have their young.

In the Arctic, gray whales need to fuel up, building the reserves necessary for such an arduous migration. Their goal is to start the trip with a full tank. But lately, that's been harder to do.

Sea ice is shrinking in the Arctic, one of the fastest-warming places on the planet. That fundamental shift is altering the ecosystem, reducing the availability of the tiny, shrimp-like animals that gray whales like to feed on. Gray whales need to eat more than a ton of them per day. As a result, scientists believe that whales are running out of fuel before they can finish their migration.

"These whales are hungry," George says. "We think they're stopping at different areas along their route to find sources of food, and San Francisco Bay has become one of those hotspots."

Heat-sensing cameras

Not far from the span of the Bay Bridge, a puff of spray, the sign of a whale breathing, appears in the distance. Shawn Henry, CEO of WhaleSpotter, pulls up an image on a laptop to show how his company's camera detects the whale's exhalation.

"That blow is a little bit warmer than the water and the air around, so it provides a very good thermal signature," Henry says.

Once it's detected, the whale's position is posted on the WhaleSafe website, run by UCSB's Benioff Ocean Science Laboratory. The Coast Guard then uses the information to alert vessels over the radio about a whale's position. Before this project, Coast Guard alerts relied on visual reports of whales from vessels during the day.

"Now with this new technology, it'll show us whales at night, so we can identify them and notify traffic," says the Coast Guard's Reed.

Another camera is also being installed on a local ferry. The Bay Area's two ferry companies say their operators either slow down or go around the areas where whales have been seen. It's tougher for larger container ships, which are much less maneuverable and are constrained to specific shipping lanes in the bay.

For now, the effort is voluntary for ships. On other parts of the California coast, researchers have seen significant compliance with voluntary speed limits from shipping fleets without mandatory regulations, McCauley says.

"I'm really optimistic that this is one of those solutions where the community comes together, and the community solves it, but we'll see," he says.

Conditions for gray whales may only get more challenging in the future, McCauley says. The whales are showing an ability to adapt, but it may only go so far.

"The world is changing, they're trying their best to change themselves," he says. "The one thing they're not doing is quitting."

Copyright 2026 NPR

Updated June 8, 2026 at 7:02 PM PDT

As vote tallies in the Los Angeles mayoral election trickled in slowly over the last week, unsubstantiated claims exploded on X that a fraudulent plot was underway to deprive the MAGA-backed former reality TV star Spencer Pratt the second-place slot to advance to the November runoff against incumbent Democrat Karen Bass.

A portion of these unfounded conspiracy theories pointed to changing betting odds for the three top candidates on prediction market sites Kalshi and Polymarket to suggest something sinister is afoot with the vote count. Some influencers supercharging such fraud claims online did so in posts sponsored by the companies themselves.

"They are actually doing it. They are counting votes until SPENCER LOSES. Someone DO SOMETHING," Trump-aligned influencer Mila Joy wrote to her half a million followers a day after the election as she reshared a Polymarket post with a graph showing that Pratt's betting odds were falling on the site.

"Is CA cheating to get Spencer Pratt out?" questioned commentator David Freeman, who posts under the handle Gunther Eagleman on X, as he shared a Kalshi post showing the odds between Pratt and progressive Democrat Nithya Raman. The Associated Press called the second-place spot for Raman on Monday afternoon after her vote share overtook Pratt's on Sunday.

At the bottom of both X posts, the words "paid partnership" appear in tiny font, a subtle reference to the millions of dollars Kalshi and Polymarket have pumped into programs that pay influencers to reshare corporate posts as a way to boost engagement.

The Los Angeles mayoral race is the clearest example yet of how prediction market posts about changing betting odds for candidates are being weaponized on X to sow doubt about the integrity of elections.

It's likely a preview of what's to come this year ahead of the midterm election. Kalshi and Polymarket are increasingly pervading ever more corners of daily life. Their rise has set off dozens of legal battles and raised novel questions about the ways betting on just about anything can have real-world consequences. Now it appears they are driving the latest battlefield in political misinformation wars on X.

"From the perspective of the influencer looking to get rich, their only job is to attract attention," Emerson Brooking, a disinformation expert at the Atlantic Council's Digital Forensic Research Lab, wrote in an email. "They will do this by sharing markets that align with what their audiences want to see. And if the betting markets are wrong, it is much wiser for them to allege fraud (and keep the lucrative promotions contract) rather than acknowledge that the gambling sites got it wrong."

In recent days, Kalshi and Polymarket have attempted to rein in some of their paid influencers. After NPR asked Kalshi about several partnership posts on Friday, the company said it told the influencers to take the posts down. Some of the posts, including Freeman's post questioning "CA cheating," have been deleted. Semafor first reported on Kalshi's crackdown.

On Monday, Polymarket told NPR it, too, is pulling back its sponsorship of some creators who were spreading election falsehoods. Joy's post is still live on X with the "paid partnership" tag, but the tag has been removed from posts by two other influencers paid by Polymarket.

"Companies shouldn't be paying people to spread misinformation," said Brendan Nyhan, a political scientist at Dartmouth College, who has reviewed the sponsored posts that flew across X. "In the Trump Republican Party, fraud allegations are going to be often received with a lot of enthusiasm, especially when people often get confused about the difference between the odds of someone winning and vote share."

Inside Kalshi and Polymarket paid partnerships 

Paying influencers as social media promoters is a type of "growth hacking" tech startups often deploy to maximize the reach of their brand in an attempt to drive more users to the services.

"It's a high-risk, high-reward situation," said Seton Hall University's Jess Rauchberg, who studies digital media culture. "But it's a strategy that gets people talking about the brand."

Kalshi and Polymarket have offered creators as much as $500 per post, according to two people who formerly worked on partnerships at Kalshi and Polymarket and who were not authorized to speak publicly about the programs.

Inside Kalshi, the approach has sparked debate over what responsibility the company has when creators promote its site by spreading misinformation and other harmful content across X, according to the former Kalshi employee.

A Kalshi spokesperson confirmed on Monday that the company now prohibits anyone in its affiliate program from questioning the integrity of an election or undermining a legal ruling or official determination about an election.

Previously, the company took a mostly hands-off approach to what its affiliate creators posted to boost one of Kalshi's markets, according to the former Kalshi employee.

Before the recent controversy, one of the only times Kalshi cut ties with a paid creator over a post promoting the company was when one of their contributors posted to X celebrating the death of conservative political activist Charlie Kirk, the former Kalshi employee said.

Similarly at Polymarket, affiliate posts were given wide latitude, as long as the person posting plugged the company's markets, according to the former Polymarket employee. And there appeared to be little vetting of creators, with Polymarket tapping former Rep. Matt Gaetz as one of its paid contributors. The U.S. House Ethics Committee found Gaetz paid an underage girl for sex.

On Monday, Polymarket said that while it does not have language specifically banning creators from posting election-related disinformation, any post denying the result of an election would violate its rules against spreading false and misleading information.

Polymarket told NPR posts from two of the creators it works with have lost the "paid partnership" tag. It has not asked creators to delete any posts, but told them about the company's content guidelines.

While the company would not specify which creators, NPR confirmed "paid partnership" tags have been removed from Jun. 4 posts by right-wing influencers Benny Johnson and Kangmin Lee sharing the same Polymarket post about Raman's rising odds on the betting site.

Seton Hall University's Rauchberg said the crackdowns are just the latest example of how the rival companies are constantly trying to one-up each other.

"They want to spread this rhetoric that 'Kalshi is for everyone, Polymarket is for everyone,'" she said. "They want to give the impression that they don't have a political affiliation, but consumers are becoming more savvy that both companies are engaging in a type of 'purity politics,' each trying to outdo the other over which is the best app to use."

Not disclosing whether a social media post was sponsored is illegal under rules the Federal Trade Commission adopted in 2024. The Trump administration has not rolled back these rules, but it has also not announced any enforcement actions.

Why California vote counting attracts fraud claims

The Los Angeles mayor race was particularly vulnerable to becoming the focus of election conspiracy theories for a number of reasons. Prediction market data may have been one of them.

Pratt, an outsider candidate who received outsized attention and engagement on X, was favored for second place on betting markets on both Kalshi and Polymarket's sites in the days before the election — even when the largest polls of likely voters showed him in third place.

While the city's mayor is a nonpartisan office, a registered Republican like Pratt faced a challenge in heavily Democratic Los Angeles. But some social media commentators cited his favorable betting odds as evidence he could reach the November runoff.

Posts about what betting markets are saying about a candidate can confuse voters who may not understand the difference between betting behavior and a poll, said Zarine Kharazian, a Ph.D candidate at the University of Washington's Center for an Informed Public, who studies online rumors related to elections.

"It runs the risk of confusing people into thinking that, 'Okay, these markets have the pulse on public sentiment about the election and who's going to win,' when that's not necessarily the case," Kharazian said.

Heading into the Jun. 2 primary, election experts were already worried that California's notoriously slow ballot count would provide the opportunity for baseless fraud allegations to blossom.

A large portion of voters in the state use mail-in ballots, a form of voting President Trump has tried to associate with fraud. Election officials must verify mailed-in and dropped off ballots, making them slower to count. The state accepts ballots that are postmarked on the day of the election that arrive within seven days.

Ballots that are counted later in the process typically skew Democratic since more voters from that party embrace voting by mail. This phenomenon has been the basis for unfounded allegations of fraud in recent years, including by Trump.

The challenge has been particularly stark this year because so many Californians waited until Election Day to drop off their ballots, said David Becker, executive director of the Center for Election Innovation and Research.

State officials are "doing what they've always done — counting all the ballots, under transparent observation from the candidates and parties, and reporting each batch as soon as they can," Becker said, "yet the profiteers and grifters are loudly echoing our foreign adversaries in spreading lies designed to delegitimize our transparent election process."

President Trump himself has claimed without evidence that there was fraud in the Los Angeles mayoral's race. He called the election race "rigged" in a Truth Social post early Monday, and wrote it was "not possible" for Pratt to lose to Raman after his initial lead when vote counting began. The first assistant U.S. Attorney for the Los Angeles area, Bill Essayli, announced on X days earlier that his office had multiple election fraud investigations underway.

Over the weekend, Essayli debunked one popular conspiracy theory circulating on X — that Pratt had received zero votes in a ballot count update — as false.

Election experts say the baseless fraud allegations in California do not bode well for the upcoming November midterm season.

"I think we're going to get punched in the face so badly on election denialism in November," said Stephen Richer, the former Republican recorder for Maricopa County, Arizona, who dealt with baseless fraud allegations in the aftermath of the 2020 election. He is now a legal fellow at the Cato Institute and a senior fellow at Harvard University's Kennedy School of Government.

Richer recalled that during the 2020 election, people trying to undermine the election results latched on to graphs that showed a blue line representing former President Joe Biden's totals suddenly jump higher as ballots were counted.

"And so now it seems that they're using these prediction market graphs to tell a similar story," Richer said.

Copyright 2026 NPR

THE GREAT NICOBAR ISLAND, India — Fireflies illuminate the edge of a forest on the Great Nicobar Island as field biologist Sumit Kumar tries to find a particularly shy creature.

A soft hoot wafts through the thicket. Kumar scans the trees with his flashlight: Sitting on a branch is a rare, wide-eyed, fat Nicobarese Scops owl. It narrows its eyes into what feels like a death-glare. Kumar smiles: "When you spot them, they look at you as if to say, 'You don't belong here.'"

And he says, they're not wrong.

The Great Nicobar Island is part of an archipelago that lies deep in the Indian Ocean. Until mainland Indians started settling here a few decades ago, its humans consisted of around a thousand indigenous folks.

It's governed by India but is so distant that it takes a flight from the mainland and a 30-hour ferry ride to arrive.

The Indian government hopes to change all that.

The upcoming Great Nicobar Project is set to transform this sleepy island into a bustling township over the next three decades.

Once complete, the island will have a civilian and military airport, a transshipment port that caters to container ships, a power plant and a new town equipped to host a million tourists a year — nearly 100 times its current population.

The project will cover an area twice the size of Manhattan, and potentially feature high rises, discos, even Disneyland-like theme parks.

Environmentalists and critics have a list of concerns. They say farms, beaches and hills will be swallowed up and a million trees will be felled. They worry about the impact on endangered animals, like leatherback turtles, largest of all sea turtles, and the Nicobarese pigeon, the closest living relative of the dodo, with its distinctive fluorescent green and orange plume.

The Great Nicobar Project "sounds like an open invitation to disaster," says Manish Chandi, a scholar who has studied the Andaman and Nicobar archipelago for over two decades. "It poses a threat to a huge amount of natural resources, its biological diversity and its indigenous communities." 

Chandi argues the purported benefits of the Great Nicobar Infrastructure project reflect a flawed understanding of "development." Residents are not the primary beneficiaries, he says. "It's a model that sees money-generation as the only way forward." He says the price of that extraction isn't taken into account.

It's a tussle mirrored in many state-backed infrastructure projects across India, from a coastal road underway in the Arabian Sea that cuts through mangrove trees to an upcoming dam in the Himalayas that will decimate chunks of forests. The clamor to protect nature has grown sharper as India sees a rise in heatwaves, glacial floods and extreme rainfall in recent years. 

After some public criticism last year, the environment minister Bhupendra Yadav insisted that the project "poses no threat to the island's tribal groups, does not come in the way of any species and does not jeopardize the eco-sensitivity of the region."

Indian ministers and departments overseeing this project did not respond to NPR's emails with a list of questions about the potential negative impacts of the project.

Why this project?

The global presence of China looms over the project.

In a press release in May, the Indian government said the goal is "to enhance India's national security, strategic and defense presence, strengthen the islands' economic position, and accelerate holistic development in the region."

And more plainly, the Hindu nationalist Bharatiya Janata Party that governs India has described the project as a "strategic gateway to crush China" in a series of social media posts.

It says the project can also help "challenge the dominance" of China in the Indian Ocean. Analysts say the shipping blockade in the Strait of Hormuz stemming from the Iran war has lent an air of urgency.

"If we think about global choke points today, especially in light of conflict in the Strait of Hormuz, India is one of many countries that are looking to secure their own supply lines," says Nitya Labh, a maritime researcher from the think-tank Chatham House.

"The project here is a great opportunity to do that because it sits along such a major international shipping route," she says, referring to the Strait of Malacca, a narrow maritime pathway that lies between Malaysia, Indonesia and Singapore.

In a 2023 press release, the Indian government said nearly 75% of India's maritime cargo today is handled at ports outside India. With a new project, it said, "Indian ports can save $200-220 million each year on transshipment cargo" and grab a share of the regional goods traffic.

There's been a massive outcry against this project for years — from former bureaucrats, the political opposition Indian National Congress party, academics and indigenous communities. They accuse the government of downplaying its ecological impact and overstating its economic and security benefits. Some have also filed lawsuits.

Others, like Abhijit Singh, a former Indian naval officer and expert on maritime affairs, have questioned the government's claims.

"This strategic and commercial gain that we are talking about seems to me a bit notional," says Singh. "But the damage to the environment is going to be very real."

Singh says India already has military infrastructure in the region to counter Chinese threats. He adds, a transshipment port only makes sense if it can lure shipping companies from their current stopovers in Singapore and Sri Lanka.

"A transshipment port does not just come up in a vacuum. It requires a logistical network. The big problem with Nicobar is that it is over 700 miles away from the Indian mainland. That means the markets and cargo production centers are quite far off from the transshipment port."

India's ruling party has bristled at criticism of the project.

In April, the country's political opposition leader Rahul Gandhi described it as the "biggest scam and gravest crime" against nature and "indigenous communities" during a visit to the island.

Days later, the ruling party accused him of sabotaging the project on behalf of China and George Soros, echoing widespread antisemitic conspiracy theories that the billionaire Jewish philanthropist seeks to subvert popular rule.

And many fear reprisals from the government for speaking out.

Nearly a dozen environmentalists, think tanks, public officials and residents declined to comment when NPR reached out, or they requested anonymity. Some said they worried about their ability to obtain funding for their projects or obtain access to the island if they publicly criticize the project.

But India's ruling party has promised the project would bring new roads, power, internet and more than 50,000 jobs to the island. The interior minister Amit Shah promised in a speech earlier this year that "in a decade, this region will draw the most tourists in the world."

For many islanders, that is a major incentive.

Two populations: Settlers and islanders 

On a recent spring afternoon, around two hundred men and women sit in neat rows at the community hall in Gandhi Nagar, a settlement built by mainland Indians when they migrated to the island five decades ago. A dozen bureaucrats had flown down for the public hearing scheduled this afternoon. They sit behind a small table, looking somber.

At the five-hour public hearing, residents ask for guarantees: jobs, houses, farmland and a hefty payout, not the pittance they say they're being offered and that the government confirms: a dollar-and-a-half per square meter of their land.

"We're no ordinary people," says an elderly man with a long white beard, who did not give his name during the public hearing. From the 1970s, he says, the government shipped hundreds of Indian citizens from the mainland to build roads and tend to farms, and to act as India's eyes and ears against Burmese poachers and foreign powers. They lived through earthquakes and diseases, staying put even when the deadly tsunami of 2004 devastated the island. "Had we run away, the Chinese flag would've fluttered on Great Nicobar," the man says. The crowd cheers.

But for the indigenous communities, the threat is existential.

Around a hundred members of the hunter-gatherer Shompen tribe live in the Great Nicobar's rainforests. The Indian government forbids outsiders from most contact with the tribe because their bodies aren't immune to modern day diseases. In the past, thousands of the indigenous Great Andamanese people living in the region died after contact with British colonizers led to an epidemic of measles and syphilis. The nonprofit conservation group Survival International, which focuses on the rights of indigenous people, says massive tourism risks contact between the island's indigenous tribe and outside visitors.

The Indian government insists that the safeguards are in place, and the rights of the Shompen will "not be affected adversely." But anthropologist Vishvajit Pandya, who interacted with the Shompen people as part of an official study in 2019, told NPR that the project's official maps he has studied include areas they're known to inhabit.

To prevent interactions with outsiders, the government's environmental impact report proposed using barbed wire to fence off areas Shompen communities are known to inhabit.

The island's other indigenous folk — the Nicobarese — are also worried. They do have contact with outsiders and have spoken to reporters, including NPR.

For generations, around a thousand Nicobarese people have lived in coastal villages. It was a simple life, says chief Barnabas Manju. "We fished in the sea, got honey from forests, squeezed oil from coconuts."

The 2004 tsunami wrecked their thatched-roof homes near the coast and forced them to relief camps in the island's administrative center. Manju says Indian officials had promised to help them return when things got better. That never happened.

Over syrupy tea and biscuits, Manju and his three deputies recalled how the lives of his community members have fundamentally changed. They now labor on building sites for money and sleep in tin sheds instead of the thatched-roof homes in their village. Their diet includes processed food. They buy fish and coconuts from the market instead of doing their own hunting and gathering for free.

Four years ago, Manju says, officials told him about the Great Nicobar project. "They had brought with them a consent letter. They didn't even give me time to read it — and just asked me to sign."

Manju says they promised him the project wouldn't impact their ancestral lands. When he saw the project's maps later, he realized part of the port would be built over his community's ancestral lands.

But Manju says what keeps them going is faith.

Every Sunday, they pray at their church, then ask for blessings for everyone: friends and family, island officials and India's prime minister.

Manju says he will lead his people back to their thatched-roof homes in their villages one day. And when that happens, he hopes officials understand why it was so important to them: "Because a country's development shouldn't come at the cost of its people's identity."

Leesha K Nair is a freelance journalist from the Andaman and Nicobar Islands, focusing on intersecting themes of environment, climate, mental health and Indigenous issues.

Copyright 2026 NPR

For new discoveries, everyday mysteries, and the science behind the headlines, follow NPR's ShortWave podcast .

Over a century ago, the German psychologist Wolfgang Köhler conducted what became a classic experiment. He suspended a banana to keep it just out of reach of a chimpanzee, placing a pile of boxes and crates nearby. The chimp soon stacked up the boxes, climbed them and grabbed the treat.

This was evidence, Köhler believed, of spontaneous problem solving by the chimpanzee; no training was required. It was the kind of thing that humans do all the time.

Since Köhler's early work, researchers have conducted similar experiments involving an out-of-reach reward and an object to stand upon in birds and elephants. And both have solved the problem successfully.

Olli Loukola, a behavioral ecologist at the University of Turku in Finland wondered whether bumblebees — short-lived creatures with miniscule brains — might be capable of the same task. And in a paper recently published in the journal Science, he and his colleagues present evidence that they are.

Untrained bumblebees consistently managed to roll a small Styrofoam ball into a position that allowed them to climb atop it to reach a rewarding stimulus overhead.

"I wasn't expecting that high success rate," Loukola says. He concludes that "very tiny brains can solve super complex problems."

Expecting greatness in the smallest of packages

After studying bumblebees for about a decade, Loukola has come to expect the unexpected. If you don't have limitations on what's possible for them, he says, "you can go wild and crazy and find completely novel stuff."

His early work proved him right. He showed that bumblebees appeared able to "learn to use tools," he says. "They learn socially from each other; they even understand the role of their partner in cooperative tasks."

Loukola has been drawn to studying tasks long considered the domain of animals with backbones. So he decided to see whether his bumblebees might be capable of a variation on Köhler's classic experiment of the banana and the box.

But he had to replicate the experiment for an organism that could easily fly to reach its reward.

Loukola, who was at the University of Oulu in Finland at the time, first trained the bees to associate a small blue circle with a sweet treat. "Bees are super fast in associating things together," he says. "They will learn immediately that blue means reward. Then they start searching for blue stuff."

He then placed just the blue circle without any sugar water on the ceiling of a hollow puck-shaped container that was about an inch high.

"We designed the arena so that it's just annoyingly [a] little bit too high for them to stand and reach the ceiling," he says, "but too tiny for them to fly."

Loukola video recorded his experiments. "With the videos, you can clearly see what is going on," he says.

In a recording of the first experiment, a bumblebee is inside the puck alongside a small Styrofoam ball. Remarkably, bee after bee in the video grabs on to the little ball and starts moving it around.

"Bumblebees, they love rolling balls," says Loukola. "Some of them needed more time and made more errors. But then they continued."

Eventually, almost three-quarters of the bees moved the ball beneath the blue dot. They then climbed atop the ball, using it like a stepstool to touch the ceiling and reach the otherwise unreachable reward.

"I planned the experiment so that it's challenging for the bees," he says. "They really need to understand the task in order to solve it."

Cognitive flexibility

There's an alternative explanation to what motivated success in that first experiment, however. Maybe the bee wasn't purposely directing the ball towards the reward.

"It's possible that the bees don't need to understand anything," Loukola allows. "Is this really goal-directed behavior or is this just playing with the balls and solving these tasks by chance?"

So in a subsequent experiment, Loukola and his colleagues introduced barriers within the arena to block the blue dot from view. The bee could no longer see the dot unless it maneuvered around the barrier. The ball was then introduced in a different part of the enclosure.

This time, some 80% of a new batch of bees rolled the ball under the blue circle, convincing Loukola that the bees had solved the problem spontaneously. It's a first, he says, for an insect with a brain the size of a sesame seed.

"We had this underlying assumption that somehow bigger brains means more powerful computations," says Cat Hobaiter, a primatologist at the University of St. Andrews who wasn't involved in the research. "And so demonstrating this in the bumblebees is really wonderful."

Hobaiter says that this study does a good job replicating similar experiments conducted on animals across the animal kingdom. "Intelligent brains come in really diverse shapes and sizes," she concludes.

The cognitive flexibility that Loukola's individual bees demonstrated may pay off in the wild when environmental conditions change suddenly and the insects must collect pollen and nectar no matter what.

"Today they might find flowers from here, but tomorrow those flowers are not blooming anymore," says Loukola. "If the workers can flexibly find new ways to get food for the colony, that's the skill that they need to have."

And, Loukola says he has all sorts of future research ideas with bumblebees. He wants to examine their body movements, microgestures and grooming behaviors to see if the insects have a tell preceding their moment of insight. One day, it may even be possible to image the bumblebee brain while it's solving a problem like the one it was presented with here.

Loukola knows that more surprises await. The bumblebee continues to impress him.

"When I started, the [cognitive] limit was somewhere here," he says, indicating a low point with his hand. "And now it's much higher."

"We have to be smarter to develop or design experimental setups where we can test their real limits," he adds.

He's not sure what those limits are, but he knows that he hasn't reached them yet.

Copyright 2026 NPR

Early results from Tuesday’s primary election show voters in El Centro and Westmoreland overwhelmingly rejected Carlos Duran’s bid for the Imperial Irrigation District Board of Directors. Instead, they voted to reelect incumbent director Alex Cardenas, who has served in the role since 2018.

As of Friday morning, Cardenas had over 1,700 votes, nearly double Duran’s total of approximately 900 votes. In a phone call, Cardenas said he saw the results as a sign that voters valued experience, ethics and transparency.

“The voters and the rate payers and water users spoke loud and clear,” Cardenas said. “They want a transparent government that doesn't placate to special interests.”

IID is the primary provider of power and water in the region. The utility delivers electricity to more than 160,000 customers throughout the Imperial and Coachella Valleys.

The agency also oversees the generations-old claims of Imperial Valley farmers to water from the Colorado River and is currently engaged in urgent talks over the river basin’s future.

Duran’s defeat was a blow for Imperial Valley Computer Manufacturing (IVCM), the Huntington Beach-based developer backing his campaign.

The company is trying to build a 950,000-square-foot artificial intelligence data center complex in the Imperial Valley. It had spent $30,000 to support Duran, a local journalist and online personality who had previously worked for the company as a spokesperson.

Cardenas, by comparison, raised around $20,000 from a mix of small donors, farmers and loans from himself.

The coalition of residents who oppose data center development in the Imperial Valley celebrated Duran’s defeat. Francisco Leal said Duran had sharply criticized IID’s current leaders but hadn’t explained his own plans.

“He drilled really hard on just bashing and talking bad about his opponent,” said Leal, a resident of the City of Imperial and a lead organizer of the coalition, known as NIMBY Imperial. “When he should have been campaigning on ways to improve and do good things for the community.”

Duran did not respond to an interview request.

The race for IID’s Division 1 seat has become one of the most prominent examples of how the nationwide data center boom and surging opposition have emerged as a driving political force in the Imperial Valley this year.

Developers such as IVCM have looked to the Valley for its energy infrastructure and availability of industrial land. Late last year, IVCM Chief Executive Officer Sebastian Rucci told KPBS he also hoped to provide tax revenue and some jobs for the region, which has one of the highest unemployment rates in California.

But IVCM needs the utility to agree to provide energy to get their project up and running. In March, Rucci told KPBS that Duran would support their plans if elected.

“In my view Carlos Duran is an excellent candidate,” Rucci said in an email. “He is not running to spare the data center of its obligations, on the contrary, he has my blessing to secure every voluntary improvement from our project.”

Duran’s candidacy, however, alarmed many Imperial Valley residents who have deep concerns about the proposed data center, which could need nearly double the amount of power that the entire county used in 2024 and 750,000 gallons of water per day.

The company has prioritized speed in its efforts to build its massive AI computing complex between the cities of Imperial and El Centro, designing it to avoid an in-depth environmental analysis.

Rucci did not respond to an interview request Thursday.

Some other utility officials said they were excited and relieved by the results. Earlier this year, IID Chairwoman Karin Eugenio said she saw Duran’s campaign as an open bid for political power in the Imperial Valley.

“I was terrified,” Eugenio told KPBS Thursday. “What that would do to the integrity of our board and how that could compromise the safety of our county.”

Eugenio, an outspoken critic of IVCM’s data center project, was also up for reelection this week. She currently holds a strong lead over her opponent, Eric Rodriguez.

The effects of artificial intelligence on learning are still largely unclear. But a new NPR/Ipsos poll of K-12 teachers found that nearly 3-in-4 believe AI has bigger implications for education than past innovations like the internet or computers.

The nationally representative poll surveyed 545 respondents and paints a complex picture of teachers' views on AI: Many are using it to save time and improve their teaching materials, but a majority of teachers are worried AI is making it harder for students to learn to think for themselves.

"We're in an environment where teachers feel like this is going to fundamentally reshape the future of education moving forward," says Mallory Newall, a senior vice president at Ipsos. "They have serious concerns about AI's impact on how they relate to their students and how students relate to each other."

And schools have a role to play: A resounding majority of polled teachers — nearly 8-in-10 — think schools should teach responsible use of AI.

"To me, that sends a very clear message that teachers are acknowledging that AI is having humongous implications on education as we know it," says Newall. "It's not going away. And so now is the time to act."

More of a teacher's helper than a classroom tool

The poll shows students aren't widely using AI in the classroom – at least not yet. A little more than half of teachers say the technology isn't being used in class by students at all, while about 2-in-5 teachers say students are using it in class at least once a week.

Meanwhile, a majority of teachers polled — 6-in-10 — say they've used AI themselves to help with work tasks.

Loading...

Michele Naber, a veteran biology teacher at El Toro High School in Orange County, Calif., says she allows her students to use AI during certain lessons to teach them how to properly prompt chatbots and verify their accuracy.

For example, she says she'll tell her students to ask ChatGPT to describe a particular animal's physical characteristics and habitat, then verify what the chatbot generated with reliable sources. She says the lesson shows students that AI still gets it wrong sometimes.

"That's one of the things that has to be taught: You can't take it literally," Naber says.

She says she's also had success using AI to generate multiple choice questions for assessments.

"That's something that normally, as a teacher, would have taken you probably upwards of an hour … and it minimized the entire task to five minutes. That's helpful."

A majority of surveyed teachers who report using AI on work-related tasks say it saves them time, but most — 63% — say that time savings equates to two hours or less per week.

Joann Purcell, a math teacher and instructional coach at Downers Grove North High School, in the Chicago suburbs, says she's found AI useful for coming up with professional development activities for her fellow educators.

But she doesn't use AI with her students. And Purcell says it isn't reliable enough to generate math questions.

"It's a pain in the butt to go through and see where the mistakes are, and I feel like if I have to do that, I might as well just write the question myself," Purcell says.

Are students learning to think for themselves?

More than half, 54%, of polled teachers say AI makes it harder for students to learn critical thinking skills.

Christa Corricelli, a special education teacher at Saugus Middle/High School outside Boston, says AI could be a valuable technology for learning, but too often students are using it as an answer machine — not a tool to bolster their thinking.

"I think students who aren't already intrinsically self-motivated to be critical thinkers, like that top 1% of the class … I think people who are not already that personality type, we're going to see those critical thinking skills atrophy over time," Corricelli says.

Loading...

Naber, in California, feels a deep responsibility to teach her students that humans must always interrogate and verify what AI generates.

"I care about [my students]. I want them to be able to look at the world and figure out things for themselves, not rely on a piece of software," she says. "If we stop questioning what it says, we can be led to believe anything. And that's what really scares me."

More than half of teachers polled — 55% — think AI is mostly just a shortcut for students to avoid doing more work.

However, Ellie Rodriguez, a special education teacher at Royal Palm Beach Community High School near Palm Beach, Fla., says AI can be especially helpful for students with disabilities. She explains that one of her students, who is on the autism spectrum, recently used AI to get help with an assignment.

"I praised him," Rodriguez says, because he wouldn't have been able to complete the assignment without the help AI provided.

"It got him to do the work, but hopefully, too, it helped him to apply using resources – like you would use an encyclopedia, like you would use a library book – to find your answers," she says.

But Rodriguez does worry the technology could hamper learning for students who don't have disabilities or who are capable of doing assignments without the assistance of AI.

And she says she and her colleagues, including some English teachers, are deeply concerned about the impact AI is having on students' ability to think for themselves.

AI is eroding student-teacher trust

Nearly 6-in-10 surveyed educators say AI is eroding the level of trust between students and teachers. About 4-in-10 say they've required more assignments to be done by hand, and 4-in-10 also say they've required more assignments to be done in class as a result of AI.

Newall, with Ipsos, says the erosion of trust caused by AI is "one of the biggest red flags in the data."

She says that issue is compounded by another survey finding: 70% of teachers believe the public's perception of them has gotten worse.

"What that tells me is that they are trying to navigate some very complex challenges in an environment that is already rife with mistrust," Newall says.

Naber, in California, has had to adapt to how easy it is now for students to fake assignments. She says for years she offered extra credit to students who participated in beach cleanups and habitat restorations outside of school. All they had to do was show her a picture to prove they were there, she says. But then Naber's son showed her how easy it is to use AI to create a fake image of a registration table for such an event.

"I had to stop doing that because I can't verify it. That was sad," she says.

Naber says she's also modified her curriculum so that all lab work is done in class in front of her, and homework matters far less for students' grades.

"Teachers are much more suspect of things that students do outside the classroom and I hear a lot of comments like, 'Well, we can't do it this way because they're just going to use AI,'" says Corricelli, near Boston.

Josh Kauffman teaches seventh-grade English at Alabama Destinations Career Academy, a virtual public school that serves students across the state. He says he's noticed a substantial uptick in the number of AI-generated assignments his students submit – and because it's a virtual school, he can't do things like require more in-class work. Instead, he says he tries to persuade his students that there's value in their own writing.

Loading...

"I tell them I would rather deal with all of your typos and know that they're yours than to wonder how much you're standing on other people's shoulders to do your work for you," Kauffman says.

Purcell, in Illinois, doesn't necessarily think AI has eroded trust. She says students found ways to cheat on assignments long before AI came along.

"I think teachers need to be creative in how they use it and force kids to think with it just like they would with any other tool," she says.

Schools aren't providing teachers with much guidance

Many educators are having to adapt to AI with little guidance from their school or district, according to the survey results. Among teachers whose school provides AI software, only 35% say they have a formal policy on teacher use of AI — meaning schools are more often providing the tools without a formal policy for their use.

About half of all polled teachers say their school hasn't offered any guidance on AI, or they're not sure what the guidance is.

"I think teachers are looking for additional guidance from their district and from their students, frankly, on what AI is going to mean for the future of education," Newall says.

Only about 4-in-10 teachers say their school offers professional development or training related to AI, according to the poll.

Rodriguez, in Florida, says she hasn't received any training on the technology, and she wishes she could.

"They need to teach us how to apply that information to what we do and most importantly to how we teach to be able to utilize [AI] in a positive way," Rodriguez says.

Kauffman agrees. He says there isn't enough attention being paid to "how to teach what we are teaching differently to account for the flexibility and the resources that AI can make available."

Corricelli isn't totally surprised by the lack of training. She says schools are often slow to adapt to change, and that's been a challenge for educators.

"I think we're all just kind of trying not to drown with the whole thing," she says.

This reporting was supported by the Omidyar Network's Reporters in Residence program.

Edited by: Nicole Cohen
Audio story produced by: Lauren Migaki

Copyright 2026 NPR

Nearly two decades ago, four graduate students from MIT united around a shared idea. "We believed that programming cells would ultimately be more important than programming computers," says Jason Kelly.

It felt like an outlandish bet at the time. Things like gene editing or testing new molecules typically demanded many hours in the laboratory — carefully mixing hundreds of chemical cocktails by hand and pipetting them into petri dishes, tasks that required an enormous amount of human labor.

The first step, they figured, was to speed that process up. So they started a company to replace those human lab workers with robots.

Early potential investors, Kelly recalls, were not excited.

" We were living on ramen, buying equipment on eBay, and we could not raise venture capital," he says of their early days running their startup.

Then came the artificial intelligence boom. In 2014, Kelly remembers reading a blog post from Sam Altman, roughly a year before he went on to found OpenAI. Kelly recalls that Atlman wrote about the potential to automate biotechnology the same way he imagined automating other kinds of technology. The two started talking.

" I was like, man, thanks for this blog post," Kelly recalls. "We've been around for five years. It is impossible to raise money."

Eventually, the Silicon Valley money started flowing.

Today Kelly runs a company, Ginkgo Bioworks, with his former classmates. It has an autonomous laboratory housed in a building overlooking the Boston harbor. Using robotics and AI, Kelly and his co-founders say that they are building the science labs of the future — where human scientists oversee robotic versions of themselves.

Pipetting robots

" Pipetting robots," Kelly says while giving a tour. "I'll show you where we do that."

Robots are arrayed around the lab, each working on separate science projects. They look nothing like humans — more like one-armed machines, each encased in glass like museum displays. A big screen at the front of the room shows a color-coded schedule of the experiments and each robot's tasks for the day. Below it a track resembling an oversized toy train set runs through the room, delivering equipment from one robot to another.

Gingko Bioworks does all kinds of work here including pharmaceutical, agricultural and government contracts. Current projects include engineering microbes for better fertilizer and creating proteins that will make snow or ice. They do a significant amount of research on pharmaceuticals.

"That one there," says Kelly, gesturing to a petri dish being ferried from one robot to another, "that has actual live cells in it."

To do this work, scientists use AI to translate experimental designs into instructions for robots about the work they need done in the lab.

Empowering the robots to be the scientists

Recently Gingko's scientists have been experimenting with taking things a step further – empowering the robot to be the scientist.

"The really, really wild moment was the first time I saw a lab notebook entry written by the model," says Reshma Shetty, another of the founders.

Shetty recently worked on a collaboration with OpenAI. Working through ChatGPT, they challenged the bot to create a certain protein. Typically, this level of thinking is left to the scientists, not unlike writing a recipe and handing it to a robot to execute. Now they were asking the bot to write the recipe for them.

"We had no idea if it would even be able to make protein," says Shetty.

The bot performed better than they expected. In comparison to human work, they concluded the protein synthesis was a 40 percent reduction in costs. It ran more than 30,000 experiments in 6 months. They've published these results, though the paper has not been peer reviewed.

Both Shetty and Kelly stress that humans are still needed to provide the right questions and constraints for experiments. Still, Shetty says it has already fundamentally changed the way she practices science.

"Normally, I rush through designing my experiment because I need to get it done so that I can actually do all the pipetting in the lab and set it all up," says Shetty. 

Now, she says, she spends more time designing her experiments so that the robot can do them for her overnight.

New access to science comes with risks

Some people warn these new freedoms bring new dangers. Drew Endy, who studies bioengineering at Stanford, says that artificial intelligence opens the door to the possibility of people with little to no training in science running experiments with questionable goals.

He and some colleagues recently wrote a report illustrating the ways artificial intelligence could be used to do things like mass-produce viruses or create other biosecurity threats.

In general, says Endy, "I'm thrilled about AI and science right now as a researcher," but he is also worried about risks including potential bioweapons programs in other countries. "I'm not excited about that."

He notes that regulations and policy to mitigate these risks are within human reach, but need to be prioritized well in advance of a biotechnological disaster or warfare.

Until now, says Endy, biotechnology has been naturally insulated from these risks through intellectual gatekeeping. " Biology has traditionally been hard for people to really gain control over," he says. "AI could nudge it a little bit more towards concentration of power."

For better or worse, Jason Kelly says he foresees a day when the practice of science is democratized.

"I do think you'll have a culture clash," says Kelly, "coming of what happens when everyday people can ask scientific questions."

Copyright 2026 NPR

I was winding down my work day here in Los Angeles when my phone rang at 5:37 p.m. from a blocked number. It was former Congressman George Santos. He was boiling with rage.

The day before, I published a story revealing that the Justice Department and the Commodity Futures Trading Commission had opened investigations into his trading activity on the prediction market site Kalshi.

Officials at the company detected that he was betting against his appearance at Trump's State of the Union address in February, just as he posted a video on X gushing to his followers about how excited he was to attend. With the help of three sources, I was able to confirm that Kalshi referred the matter to federal authorities in the Southern District of New York and Washington.

Before the story, I emailed him, and he called me from a blocked number. So when my phone buzzed again from a blocked number, I had a pretty good sense of who it might be.

Santos, whose political rise and fall was characterized by a notorious trail of lies and falsehoods, claimed my story was riddled with errors. He said "my lawyers have been calling the Department of Justice all day, and they can't find any investigation."

As we were talking, I asked if I could record the call. He said no. I was in front of a keyboard, though, furiously jotting down every word.

I asked him who his lawyers are, and he refused to answer. I questioned whether he really does have attorneys. He replied: "I'm George f*cking Santos, of course I have a legal team."

He then proceeded to name-call and attack the reputation of NPR, the kind of invective that's common when reporting on people who try to discredit reporters and news organizations for stories they don't like.

What Santos said next took me aback, even by his outlandish and brazen standards.

"This story is going to get you a gun in your face," Santos said.

I asked him what he meant by that.

"You know what I mean."

It did not exactly feel like an imminent threat to my life that a convicted fraudster expelled from Congress who lives thousands of miles away from me in Pennsylvania's Pocono Mountains was lodging violent words at me.

It felt more bizarre than threatening, but then it grew even stranger and more confusing.

While he had been calling me from a blocked number, I tracked down his cell phone from a public records search, and I sent him a text message confirming it was him. He greeted it by launching into full-throated denial. "I NEVER SAID 'this story would get a gun in your face, I said 'it'd blow up in your face," he wrote in a text.

He then called me "an insane person," "a clown," among other broadsides.

Santos was front-running his denial of his violent threat before I ever publicly confronted him for doing it.

Not long after, he took to his favorite megaphone, X, to tell the world this never happened, claiming I "was now making things up," even though I had yet to publicly reveal what he had said.

"I've interacted with hundreds of reporters in my life… not once was I ever threatening or aggressive… sassy? Sure but aggressive and threatening? NEVER!" he wrote.

In the post, he added: "He's now demanding I disclose the names of my lawyers 'or else' (only God knows what that means.)"

Which is a fiction. I did ask him who his lawyers are, but the "or else" is Santos fantasy, perhaps his way of turning me into the menacing actor in all of this.

Threats against journalists are unnervingly common. Most are flung by online trolls or aggressive attorneys and PR handlers trying to protect their clients.

At times, it can be hard to know when it's frivolous or hyperbole, or when you should take it seriously.

I grappled with whether I should call this out at all. After all, drama and attention are oxygen to someone like Santos, who has become something of an internet troll since President Trump commuted his prison sentence, giving him his freedom back and his access to X.

But given his mischaracterizations of how it at all went down, and his quick handiwork trying to cover his tracks, I thought it was worth setting the record straight.

Since I published the story on the federal investigations into Santos, the Associated Press reported that Polymarket has cut ties with Santos, who had been paid by the company, a rival to Kalshi, to boost social media posts featuring some of its prediction markets.

He also appears to be hoping to drum up some new business on Cameo, the site where celebrities are paid to record personalized videos.

Santos is now offering 55% off his Cameo videos, meaning for $150, he will record himself saying just about anything. Although there's no guarantee he won't later deny it on X.

Copyright 2026 NPR

Stay up to date with our Up First newsletter, sent every weekday morning.

Kirby Plessas doesn't have an AI subscription. She has two.

As a self-described technophile, she uses chatbots outside of work to plan family parties, tweak cocktail recipes and once to diagnose a broken wine cooler's motherboard. All the help Plessas gets from AI justifies the $40 a month she pays for both OpenAI's ChatGPT and Google's Gemini.

"I wouldn't doubt that within a year I'll probably have a Claude subscription as well," she said.

But for most people, she believes, free AI is good enough. And judging by their spending, Americans agree.

Only about 3% of households were paying for AI in February, using the most recent numbers available from the Bank of America Institute, which researches consumer trends based on the bank's customer transactions.

Yet while the number of personal subscribers remains small — plenty of workplaces and universities do pay for AI services — their ranks are growing fast. About 10% more households paid up in February compared to a year earlier, according to the institute.

"If you think back to Netflix and streaming services, at the beginning the growth was quite slow," said Sekoul Krastev, cofounder of the Decision Lab, a research firm with an emphasis on behavioral science. Krasteve said it isn't the norm to pay for AI — yet.

"Once that status quo is created, subscriptions will definitely start to go up sharply just the way we saw with streaming services," he said.

Plessas puts it this way: "There's a thought out there that we're all going to get addicted to using AI. So when the free ones go away, everyone will have to pay."

Subscriptions get you more 

Even if they worry about how it will affect their daily lives, most Americans use AI. In fact, 51% of Americans said they use AI to research what they're curious about, according to a Quinnipiac poll released in March.

They're just using free versions. Most AI platforms offer them — to a point. For example, OpenAI lets users send its default model 10 messages every five hours before booting them to a weaker version.

"You would get a little notification that says 'Your time has run out,'" said Pam Dean, who now subscribes to both ChatGPT and Claude. "You're stuck in the middle of something and then you couldn't continue."

The main selling point for subscribers is that they get to ask smarter models more questions, more often. A $20 monthly ChatGPT Plus subscription lets users send up to 1,280 messages a day before a downgrade.

Subscribers also get more advanced features, such as being able to create tailored versions of ChatGPT to fit their specific needs.

To prepare for a possible move from Los Angeles to Mexico, Jim Arnold created "Francisco," a ChatGPT version that acts as his Spanish tutor. Francisco corrects Arnold's Spanish and has conversations with him through his computer speakers. Arnold fine-tuned Francisco to speak slower, and to take longer pauses before replying. That gives Arnold a chance to think of the right Spanish word before Francisco interrupts.

Still, Arnold never tried the free version of ChatGPT, and was wondering if paying was worthwhile. So he asked the AI. "And of course, you know, ChatGPT's answer was that I should keep what I'm doing for sure," Arnold said.

Today's version of ChatGPT will remain free …

OpenAI has about 50 million subscribers. Nick Turley, the head of ChatGPT at OpenAI, believes there's room to quadruple the number of users who pay.

Still, Turley said, subscribers will always be a minority of the people who use ChatGPT. He said the company's mission is to keep this tech accessible to benefit everyone: "And one of the ways that we do that is by making sure people can access it for free."

The birth of ChatGPT subscriptions wasn't about revenue, but keeping the site usable, Turley said. AI takes expensive computing power, and in the early days after ChatGPT's launch, putting up a subscription paywall helped the company keep up with demand while still allowing deeper access to those willing to pay up.

"We weren't even trying to make money," Turley said about the launch of the subscriptions. But, he added, "It's been a really nice side effect that that also generates revenue and builds an amazing business."

…but you might have to pay for the next version

Turley believes OpenAI will likely always offer a free version of ChatGPT. But he is not promising a free version of its eventual successor.

That will likely be an agentic AI, a type of AI that can act on its own to do tasks for the user or an organization. Think of an AI agent that can plan a vacation for you, or do your shopping.

OpenAI already offers agentic services, but Turley envisions a version that acts like a personal assistant, thinking of ways to improve your life while you're away from the screen. And that assistant could be so advanced and expensive to run that it might need to stay behind the paywall.

"If we do our job right, many people are going to want to pay and subscribe," Turley said.

Could AI companies use advertisements? 

AI companies do have an alternative option to cover the cost of their services — advertisements. That's how many of the biggest websites, like Facebook, have kept their services free for years.

But Jeff Hancock, head of Stanford University's Tech and Impact Policy Center, said there's a reason AI companies should be cautious about adopting an ad model: "People hate it."

Ads incentivize social media companies to make sure users keep scrolling. The more screen time, the more ads get viewed. And endless scrolling has caused concerns about people's mental health and attention spans.

"People just don't like how social media fits into our lives now," Hancock said.

The second reason to be careful about adopting ads, he said, is that AI is a different kind of tech from social media platforms. The kind of AI personal assistant that Turley described would ideally work in the background, so the user would actually spend less time looking at the screen — a poor match for a business model where the main revenue comes from seeing ads.

"AI platforms have the potential to create a completely different economic model by the incentive being: Is this useful for you?" Hancock said.

Sarah Womer subscribes to several AI platforms. In fact, she rotates subscriptions across different AI, saying they each have their own benefits and personalities. "You're not limited to one flavor of AI, just like you're not limited to one flavor of ice cream," Womer said.

She prefers using OpenAI when she books vacations. But Open AI has been piloting ads for free users and Go tier subscribers, and Womer worries AI companies could skew models to favor their advertisers.

In a statement from OpenAI when the ad test was launched in February, the company promised that ads would be clearly labeled and that they would not influence ChatGPT's answers. Users would be matched with ads, the statement said, based on "the topic of your conversation, your past chats, and past interactions with ads." But promises like that haven't put Womer at ease. So when she uses AI to research products she wants to try — like the best red light face mask — she instead uses her subscription to Kagi, an AI and search engine platform that boasts about its privacy standards and lack of ads.

AI could be paid for through bundling 

ChatGPT launched less than four years ago, and the revenue model for AI services is still very much in flux.

The global investment advisor company BlackRock held a summit in March about upgrading the country's infrastructure, where OpenAI CEO Sam Altman floated the idea that AI services could be sold like electricity – essentially billing customers based on how much they use.

Greg Portell, lead partner at the Kearney consulting firm, believes another potential option is for companies to bundle AI in with other subscriptions — like Amazon Prime or internet service.

He thinks this is likely since Americans are already fatigued by subscriptions. Rather than get consumers to sign up for another monthly expense, it would be easier to just add it to one they're already paying for.

"Every consumer says they have too many subscriptions, but yet every subscription provider is looking at how they can bundle things," Portell said.

Of course, many Americans already use a paid version of AI at work. Those corporate accounts often come with limits on personal use, but Portell wonders if in the future workplaces will lift those limits. Maybe having your job pay for your personal access to Claude or Gemini is the new employee perk.

Google is a financial supporter of NPR.

Copyright 2026 NPR

The Trump administration is pursuing a bureaucratic rule change that could allow for greater political influence over billions of dollars in federal research grants. The new rule would have a broad impact on research fields, including housing and transportation. Health and science funding would be most significantly affected.

"Although research has bipartisan support in the US Congress, and trust in science is above 75% across the country, the Trump administration seems as determined as ever to mortally wound the nation's scientific enterprise," Holden Thorp, editor of Science magazine, wrote in an editorial about the proposal.

Published in the Federal Register on May 29, experts say the proposed changes would both codify the administration's strategies to dismantle certain fields of study in the U.S. and lend it new authority to "advance the President's policy priorities."

In science, the impacts could reverberate across fields of research as varied as public health, vaccine testing, biotechnology, social and behavioral science and climate science.

The proposal is animating advocacy and science groups across the country.

"This would be the end of American science as we know it," said Cole Donovan, a policy analyst from the group Stand up for Science who has been organizing to protest the change. "We're gonna make sure that it doesn't fade quietly into the night."

End of peer review as the gold standard

Since the post-World War II period, the U.S. scientific community has relied heavily on a system of peer review to offer feedback on studies and maintain integrity in research. The same has been true for federal science agencies when evaluating proposals for research funding. Typically, agencies adopt recommendations from independent advisory committees on issues including vaccine schedules, environmental standards, or census methodology.

While not legally binding, peer review in practice has been enormously influential and become part of the norms of government.

" While it's been true that peer review panels have always been treated as advisory by agencies, it was usually the combination of peer review with a non-political career expert at an agency that made the determination of whether to issue an award or not," said Donovan.

Under the new rule, peer review would not be eliminated, but political appointees — not necessarily scientists — would be required to review grants before awards are made. Critics say that effectively gives political officials veto power over projects, even when they have passed scientific peer review.

The reaction from scientists and advocates has been swift and vehement.

" If this sort of rulemaking or rule-breaking becomes the norm, then government itself will cease to work," historian Tim Snyder said in an online forum Tuesday organized by Stand Up for Science. Snyder studies the former Soviet Union and remarked that the Trump administration's proposed rule change reminded him of "late Stalinism."  

"We're asking ourselves whether we wanna repeat that Stalinist situation where people who know nothing about science are the ones who are making the decisions about what's going to go forward," said Snyder.

Others compared the moment to a period in the United States in the 1950s when government officials scrutinized people's ideology and politics rather than their credentials.

"Proposed rule would replace scientific merit with McCarthy era politics," read the headline from a statement by the Infectious Diseases Society of America.

The rule change was proposed by the White House's Office of Management and Budget, which is led by Director Russell Vought. Administration officials argue that the rule change is in the interest of efficiency. In a statement, an OMB spokesperson wrote it would "improve the ability of agencies to identify and respond to waste, fraud, and abuse."

Critics say there's no evidence that the existing peer-review system needs such sweeping changes and that empowering political actors to make decisions about science is dangerous.  

"When we're designing a study to a new cancer therapeutic, do you want Russell Vought — who is not a scientist — to determine which immunotherapy is ready to go into a phase three trial?" said Elizabeth Ginexi, a former staffer at the National Institutes of Health who has been speaking out against the proposed changes.

In addition to giving the administration increased authority, the proposed rule officially bans research on diversity, equity and inclusion or gender as grant conditions, and places broad prohibition on international scientific collaborations.

" We are involved in a huge amount of international collaboration," said Donovan from Stand Up for Science. "Much of the work that's considered high-impact is based on international collaboration."

Limited role for congressional pushback

Several Democratic lawmakers spoke at the meeting on Tuesday in an effort to rally support from the public.

" When promising research is denied because it doesn't fit the political agenda of the moment, the American people will pay the price," said U.S. Rep. James Walkinshaw of Virginia.

" The question isn't whether politics will influence research under this proposal," the Democratic lawmaker added. "That's the point."

Beyond exhortations to action from the public, however, Congress is unlikely to take action on the rule change.

The proposed rule is open for public comment until July 13. After that, OMB will review those comments before deciding whether to issue a final version.

In the event that it is passed, Donovan said, he anticipates it will "almost certainly" be challenged in court.

Copyright 2026 NPR

Stay up to date with our Up First newsletter sent every weekday morning.

HAYWARD, Calif. – In an industrial yard off a highway east of San Francisco, AT&T workers crowd around cold, hard evidence of a growing problem.

"Sitting here [is] a truck full of what is stolen copper cable," says Todd Swenson, from AT&T's construction and engineering division. The jumble of cables and wires, about the size of a truck tire, was recovered from a metal recycler. Swenson says that cable actually belongs to AT&T, and was cut down from telephone poles by thieves.

Over the past few years, there has been an alarming rise in copper wire theft in the United States and beyond. The value of copper has roughly doubled in the past year, thanks in part to increasing demand for the metal. So thieves strip it from phone lines, as well as from other infrastructure like streetlamps and EV chargers. Repairs cost companies and communities, vex corporate executives and politicians and tax work crews.

Swenson says record-high prices of copper — buoyed, in part, by the artificial intelligence data center boom — are to blame. "The higher the price of copper is at a recycler and on the market, our theft goes up. Direct correlation there," he says.

This pile of wires might fetch a few hundred dollars at a recycler. But Swenson says the damage could cost the company tens of thousands of dollars to repair.

AT&T executives have grown frustrated with the problem, which is why they've invited NPR on a ride-along to see it firsthand. So we caravan to a railroad crossing, where at 3:40 in the morning, an alarm had gone off alerting the company that cables had been cut down from nearby telephone poles.

The company suspected the work of thieves.

"What they typically do is they cut the cable down, they'll pull it to a location and they start working on stripping it," says Scott Gonzaga, also with AT&T. "Then they burn it to get the sheath off and to get it down to the bare copper." That metal is sold to a middleman or directly to a recycler.

We arrive at the tracks a little before noon, and the aftermath of the crime comes into focus.

The phone and internet cables that should have stretched over the track are gone. A pair of bolt cutters lies in the dirt, along with bits of rubber and plastic stripped from the wires. Down a trail by the train line, there's a string of freshly charred spots in the dirt — what the AT&T crew call "burn pits." An abandoned tent overflows with clothes and junk, and behind it lies a small pile of wires.

In the distance, down the path, there's something else: people.

There are two of them, about 100 yards away, pushing what looks like a large plastic laundry bin.

"This is a first," says Michael Riensch, who oversees efforts to combat copper wire theft at AT&T's global security unit. "We just happened to be here while subjects were on scene."

Gonzaga says he saw one of the two people by a burn pit when he showed up.

"We tried to follow him a little bit just to get a picture of him," Gonzaga says.

The two people with the cart don't seem to be fleeing, exactly. But they are staying out of reach of the AT&T crew. And the crew is keeping its distance, too.

They've already called the police, who are on their way.

Locked-down manhole lids

Last year, AT&T recorded more than 10,400 incidents of copper wire theft nationwide — about 200 a week. Some 4,300 of those were in California.

It's an outsized problem for the company, since only about 3% of AT&T's customers are still connected by copper wire networks. That includes households without access to cell service or fiber optic connections, some in rural areas, as well as some businesses that still rely on old wires for internet or fax connections.

Copper cables, pioneered over 180 years ago by Samuel Morse, often hang next to their modern counterparts, fiber optic lines. When thieves cut cables for copper, they often slice fiber cables, too, because they look similar. A snipped fiber cable is what tripped this alarm.

Susan Santana, president of AT&T West, is on the ride-along. She says homes, hospitals, airports, schools and more can lose connections when cables are cut. The problem is "not an easy one to solve, by any means," she says, but AT&T is trying.

"We have locked down manhole lids with extra bolts. We've put sensors across our lines. In some instances we've had to hire private security guards," Santana says.

AT&T has also offered a $20,000 bounty for information leading to the arrest and conviction of people involved in copper cable thefts.

The problem is not limited to telephone wires. Thieves have been cutting electric cables, too. The California city of San Jose has an online tracker for streetlights that have gone dark after being hit.

EV chargers are also targets. Rick Wilmer, CEO of ChargePoint, the largest charging network in the United States, says it's a problem they deal with every day. He says he got so frustrated that he started prototyping solutions on his own.

"I was so motivated I literally was going down to Home Depot and buying all kinds of different wire and Kevlar and stuff, and wrapping cables and taping it down and trying to cut it with my own pruning shears to see if it was, you know, making it any more difficult," he says.

He handed the project off to company engineers, who developed charger cables that are impregnated with cut-resistant material. The idea, he says, is that a thief might be able to hack off one of those wires, but their shears will be damaged in the process. They won't be able to hit multiple chargers in one go.

Sharing intel, engaging in surveillance

Back at the railroad crossing, the police have arrived. They've sat the two people with the cart down on the track's embankment and are questioning them.

In California, copper thieves face up to three years in prison and hefty fines. The state recently tightened regulations on scrap metal dealers, and bills in the state legislature would raise the penalties for organized copper wire theft and individuals holding wire without proof of ownership.

Rob Bonta, the state's attorney general, says enforcement is key. "The science shows that if people think they will be caught for committing a crime, then that deters them from committing a crime. Even if the punishment is high, but they don't think they're going to get caught, they'll still commit the crime," he says.

He says authorities have been using a playbook of cross-jurisdiction collaboration that previously helped the state tamp down catalytic converter theft and a rash of smash-and-grab robberies at retail stores.

"We started teaming up and engaging [in] investigations, sharing intel, engaging in surveillance. A lot of what worked there is working here," he says.

"So, more participation, more resources, more focus, more dedication, more prioritization by law enforcement agencies, who have a lot to prioritize," he adds. "These are folks that are tackling — you know, murders and rapes and robberies and other important crimes, too."

A frustrating pattern

AT&T also has a longer-term solution in mind: The company wants to move on from its aging copper wire networks, shifting all of its phone service to fiber optics.

But California law won't let it.

Once a monopoly, AT&T is considered the state's "carrier of last resort." Santana says that means it is required to keep copper running in order to serve that tiny fraction of customers who still use it — despite a range of new technologies and offerings from competitors.

"In an ideal world, California would find a way to help us transition, would identify a pathway to transition from maintaining a copper legacy network to allowing us to invest our resources into fiber and wireless, which is what most of the other states in this country have already done," she says.

In late May, AT&T filed a lawsuit against California to allow it to discontinue its legacy copper wire-based phone service in line with Federal Communications Commission authorization to phase it out. The company has pledged $19 billion to modernize the network through 2030.

Bonta's office says it is reviewing the complaint and will respond appropriately in court.

Not everyone is in favor of AT&T retiring its legacy copper wire network, including consumer, farm and small business advocacy groups who argue that doing so could deprive communities without other reliable options of a critical means of communication — especially in an emergency.

At the railroad crossing, Rommel Maghonay, a splicing manager, is overseeing part of the work to replace the sliced cables. It's the fifth or sixth cut cable due to suspected theft that he says he has had to deal with in the past three days.

"On a typical day I've got to pull two, three splicers off their normal assignment" to fix cut lines, he says. "It sets us back on our normal work."

This job will cost AT&T crews four to five hours, he thinks. "It's frustrating sometimes," he says, "because it's happening so often."

The police, meanwhile, finished questioning the two people. With no probable cause — no witnesses to the actual cutting of the cables, and no tools or stolen property on the people — they let them go.

The Hayward Police Department declined to make records relating to the case available to NPR, saying it is still open and under investigation.

Copyright 2026 NPR

Scientists who've spent decades learning how the brain works say they're now ready to start fixing it when it breaks.

That's the premise of the Brain Health accelerator, a collaborative effort launched by the Allen Institute in Seattle, which has become a major player in brain research.

The $400 million initiative includes plans to develop new genetic therapies — a term that includes gene editing as well as traditional gene therapy — for diseases including Alzheimer's, Parkinson's, ALS, Lewy body dementia and Huntington's.

"The latest genetic treatments allow scientists to control the activity of particular genes," says Ed Lein, who directs the institute's brain health programs. "That opens up the possibility for very specific precision therapies for brain disorders."

The accelerator is an outgrowth of the BRAIN Initiative, an ambitious research program unveiled by President Barack Obama in 2013. The goal of this public-private partnership was to create tools that would allow scientists to see the brain's inner workings and, eventually, to develop treatments.

But the effort has progressed far faster than many scientists expected.

"I am shocked at how far we've come in the last 10, 12 years," says John Ngai, a senior investigator at the National Institutes of Health who directs the BRAIN Initiative. "It's just been beyond my wildest imagination — and I've been accused of having a pretty good imagination."

Support for the accelerator includes a $200 million commitment from the Allen Institute, $100 million from the Bezos family and $100 million from sources including the National Institutes of Health, Amazon Web Services and EverythingALS. Collaborators include a wide range of hospitals, universities and research centers from around the world.

From mice to humans

The Allen Institute's Brain Health accelerator is attracting a range of scientists intent on translating their knowledge about the brain into treatments for its disorders.

Jeff Carroll, for example, was a teenager when he learned that his mother had Huntington's disease, a fatal, inherited disorder that destroys brain cells.

"The whole reason I'm in science started with this frustration with not being able to understand what was happening with my mom," he says.

Carroll eventually learned that he, too, carried the Huntington's gene.

He spent years at the University of Washington studying mice with the condition, which causes nerve cells to produce toxic levels of a certain protein.

To Carroll, the solution seemed obvious: "Since we know that all the bad stuff in Huntington's comes from this one gene … let's get rid of that gene."

But finding a way to do that was beyond the reach of his own relatively small lab at a university. So Carroll joined the accelerator effort.

"It's difficult to do the scale of research that you need with a team of five or six or even 10 people," he says. "The hundreds of people they have here at the Allen Institute [allow for] an entirely different approach to science."

And one of the first targets for that approach will be Huntington's disease.

Carroll says he's optimistic.

He notes that genetic therapies have already had success treating at least one nerve disorder, a rare genetic condition called spinal muscular atrophy. It's caused by a gene mutation that leads to the death of motor neurons in the spinal cord.

"Every kid with this horrible mutation died when they were like 18 months [old]," Carroll says, "and now they're going to high school. So things that were unimaginable can change."

Carroll says his goal is to help accelerate that change.

Big science, shared

The Allen Institute was founded in 2003 by the late Paul Allen of Microsoft fame and his younger sister, Jody Allen.

Its goal was to speed up science research. And it has done so by developing technologies that allow scientists to rapidly characterize and map millions of cells.

"We now have a complete description of the types of cells that make up the brain, and also the genetic underpinnings of their properties," Lein says. "This foundation then lets you study disease."

Allen scientists have already begun studying how Alzheimer's disease changes nerve cells in the brain.

"It affects very specific types of neurons that are lost early in the disease and then over the course of the disease," Lein says.

Genetic therapies designed to protect these neurons might delay or prevent symptoms, he says. Similar approaches might also work to preserve the neurons affected by Parkinson's or ALS.

Ultimately, though, the Allen Institute may not be the place that develops these therapies. Its policy of making its databases available to anyone means that scientists around the globe can help look for answers.

Copyright 2026 NPR

In February, four months after being released from federal prison, former Republican congressman George Santos took to social media to express his enthusiasm about attending President Trump's upcoming State of the Union address.

"I'm going to be there for the State of Union in the gallery, guys," Santos said in a video he posted to X a day before the president's remarks.

At the time, traders on the prediction market site Kalshi were placing millions of dollars worth of bets on who would attend. Santos' video confirming his presence sent odds soaring.

But he didn't show up.

"Watching SOTU from an airport tv was not part of the plan! FML," Santos wrote on X, using slang for a more coarse way of saying, "screw my life."

He posted the message as Trump was speaking, making those same odds in the Kalshi market plummet.

What Santos didn't say was that he had already placed bets on Kalshi that he was not going to appear at the State of the Union address, according to three people with direct knowledge of his trades who were not authorized to speak publicly. They say Santos misled the public and turned a profit based on that deception in the tens of thousands of dollars.

Kalshi detected Santos' trades, froze his account and referred the case to the Commodity Futures Trading Commission and the Department of Justice, which both opened investigations into Santos, according to two people familiar with Kalshi's investigation who was not authorized to speak publicly.

Neither the CFTC nor the Justice Department returned requests for comment. Kalshi declined to comment.

Reached by NPR, Santos said, "Well, that's news to me," when asked about the insider trading probe underway into his activity on Kalshi.

Santos said, "I'm not saying yes, I'm not saying no" when NPR questioned whether he had a Kalshi account.

He went on to say the co-founder of Kalshi, Luana Lopes Lara, is "a fellow Brazilian" whom he personally knows. He said he would call her to get to the bottom of whether an investigation had been launched.

Santos promised to update NPR on how the call went. He did not respond to NPR's follow-up messages.

The person familiar with Kalshi's investigation said Santos, the son of Brazilian immigrants, does not know Lara, a Brazilian-American.

Kalshi has reached out to Santos to interview him as part of the investigation, but he has dodged those requests, according to that same person.

In a post on X on Thursday, Santos said "my legal team is in contact with the DOJ to see what is going on," calling the insider trading accusation "preposterous."

Santos wrote: "I look forward to supplying any information asked of me to any agency that inquires."

A 'manufactured' political rise

Santos' trading on his own movements follows one of the most dramatic downfalls in recent political history.

The 37-year-old former Republican congressman from New York ran for office telling a series of lies and fabrications about his personal life. Those include that he graduated from Baruch College in the top 1% of his class. He also said that his mother was working in the South Tower of the World Trade Center during the Sept. 11 attacks. The stories about his academic performance and his mother turned out to be false.

He said he was Jewish. He is not. And he claimed that his grandparents escaped the Holocaust. They did not.

"Virtually everything that he put out about himself when he was running for office was manufactured," said Jonathan Entin, a law professor at Case Western Reserve University. "You had to wonder whether anything he ever said about himself had any basis in reality."

Santos' penchant for deception eventually caught up with him.

In January 2023, Santos was sworn in as a U.S. representative and by May of that year, he was hit with an indictment. Federal prosecutors charged him with 13 counts of wire fraud, money laundering and stealing money from political donors. He was expelled from Congress in December 2023. A judge later handed him a sentence of more than seven years in federal prison.

But he served only four months because Trump commuted his sentence last October.

"George Santos was somewhat of a 'rogue,' but there are many rogues throughout our Country that aren't forced to serve seven years in prison," the president wrote on TruthSocial at the time.

Prediction markets under scrutiny

The investigations into Santos' trades come during a moment of intense scrutiny of prediction markets, such as Kalshi and Polymarket, which allow bettors to put money down on everything from what will be the most expensive art work sold at auction this year to the timing of the next military strike in the Middle East.

Lawmakers in Washington have been particularly concerned about how people with insider information could manipulate the markets and defraud other bettors. In April, federal prosecutors criminally charged a U.S. Army Special Forces soldier accused of making more than $400,000 betting on the capture of Venezuelan leader Nicolás Maduro.

Last week, a Google employee was charged by the Justice Department for making more than $1 million correctly betting on search trends based on confidential information from the tech company. Both traders were using Polymarket, which is based overseas in Panama.

Traders on the messaging forum Discord suspected Santos had been duping Kalshi traders.

During the State of the Union address in February, one user with the name Roastinator.exe wrote after losing money betting that Santos would be at Trump's speech: "I want George Santos in Kalshi prison for 15 years!!!"

Another user with the handle Nate Meininger wrote: "that's gotta be a new low if you get rugpulled by George Santos," using finance lingo for when a prediction market suddenly collapses.

While Santos' account has been frozen, his name is still popping up in the world of Kalshi.

Last month, traders bet nearly $90,000 on what words Santos would utter during an interview on Newsmax.

He did say the word "rumor," but avoided the word "corruption."

Copyright 2026 NPR

President Trump signed a long-awaited executive order on Tuesday that aims to mitigate security threats posed by artificial intelligence, in a shift from the administration's hands-off approach to the technology.

The order asks AI companies to voluntarily submit their most powerful models for the government to test up to 30 days before releasing them to the public.

It also directs federal agencies to develop benchmarks to assess AI models' cyber capabilities, to create an "an AI cybersecurity clearinghouse" to review and share information on vulnerabilities, and to shore up the government's security defenses.

"Advanced AI capabilities make our Nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies," the new executive order says. "As these capabilities evolve, my Administration will continue to work closely with industry to ensure that the best and most secure technology is deployed rapidly to confront any and all threats to our country."

The executive order was expected to come out last month, but the White House scrapped signing plans over concerns that it would interfere with AI innovation. Trump said at the time he worried the order would stifle American companies' lead in the global race amid competitive pressure from China.

That earlier version gave the government up to 90 days to review advanced models before release — a timeline that was cut to 30 days in the final order.

The Trump administration has been divided over how to approach AI. While the Biden White House pushed for federal oversight of the emerging technology, Trump has sought to minimize regulation, including at the state level, even as concerns over safety risks have proliferated.

But recently, the development of more powerful AI models has spooked some federal officials, prompting the White House to reverse course and back some safety measures.

In particular, Anthropic's announcement in April that it was limiting the release of its new Mythos Preview model because of its ability to identify and exploit software security vulnerabilities set off alarm bells across Silicon Valley and Washington.

Notably, the order relies on voluntary cooperation from the tech companies leading AI development, such as OpenAI, Anthropic, and Google.

Referring to the voluntary testing framework, the order states that "nothing in this section shall be construed to authorize the creation of a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models."

Any regulations around the technology would have to come from Congress.

Copyright 2026 NPR

Reporting for this story comes from the series Safe to Drink, a podcast that follows stories of water contamination in New England by Mara Hoplamazian from New Hampshire Public Radio.

"Forever chemicals" are everywhere — in our drinking water, in our food and in products like nonstick frying pans, raincoats and even some types of floss.

Also known as per- and polyfluoroalkyl substances, or "PFAS," these chemicals have been linked to negative health outcomes, including certain cancers.

PFAS are strong and can repel both water and oil, which is why they're widely used in manufacturing. But they don't break down easily in the environment, and they can linger in our bodies. In fact, according to the Agency for Toxic Substances and Disease Registry, nearly everyone in the United States has PFAS in their blood.

"They have really unique properties, and that's part of what has made them so attractive to industry. Unfortunately, we're learning that it is also something that makes them toxic and causes problems in the human body," says Megan Romano, an epidemiologist at Dartmouth who studies PFAS.

Eliminating all sources of PFAS in your life would be impossible, according to one report from the National Academies of Sciences, Engineering and Medicine. And researchers are still trying to pin down which actions reduce exposure.

But it's worth trying. Research has found that when people remove their exposures, the levels of these contaminants in their bodies decrease over the course of several years.

If you'd like to reduce your exposure to PFAS, there are a few actions you can take. Some are simple, like looking for third-party-tested PFAS-free products. Others aren't as straightforward; the U.S. is still trying to figure out where these chemicals are showing up in food and water.

Check if your water has been tested 

If your community has water contaminated by PFAS chemicals, drinking water could be your main source of exposure. According to the U.S. Geological Survey, they're in nearly half of the nation's tap water.

Many cities and towns have already tested public water for these chemicals, so a good first step is to check with your water utility to see if they have published those results. To do that, you can call your utility's customer service line or look online to see if they've published PFAS data in water quality reports.

In many states, environmental regulators may also be able to share information about PFAS levels in public water. The deadline for water utilities to test for PFAS chemicals under the current Environmental Protection Agency regulations is 2027, so if results aren't available yet, they should be by then.

If you're on a private well instead of on public water, you may need to test your water yourself. Some states have programs to help with the expense and logistics.

The U.S. Geological Survey has a map that shows their testing results across the country. An interactive EPA map shows testing results for public water systems, with systems testing above national limits in brown. Another map, from a nonprofit called the Environmental Working Group, shows results from several national testing efforts.

Once you figure out the levels of PFAS chemicals in your water, you can compare them to the EPA's regulations. You can also enter them into the Centers for Disease Control and Prevention's blood level estimation tool, which can give you a sense of what the levels of these chemicals in your blood might be.

See if your community is installing filters, or consider getting your own

If you find out your water has levels of PFAS chemicals above the EPA's limits, check to see if your community is planning to install a filtration system that can catch those substances. Public water systems are required to reduce PFAS in water by 2029, or by 2031 under an extension proposed by the Trump administration. 

There are a few kinds of treatment systems that work on PFAS contamination: activated carbon filters, ion exchange treatment and reverse osmosis.

If you decide to get your own filter, make sure it is certified to treat PFAS contamination. You'll want to look for certifications from NSF International and the American National Standards Institute (ANSI). Specifically, filters that are "NSF/ANSI 53"- or "NSF/ANSI 58"-certified should work to reduce PFAS chemicals — though their current requirements aren't based on the latest EPA limits.

The upfront cost of in-home water treatment can range from around $20 for pitcher-style filters to hundreds or thousands of dollars for whole-home systems. But make sure to factor in maintenance costs, too — it's important to follow the instructions that come with the treatment system and replace parts as needed.

Consider your food 

When PFAS chemicals get into the environment, they can make their way into our food.

According to the National Academies report, diet is likely the primary way people are exposed to PFAS in places without drinking water contamination. But, researchers say, longer-term studies are needed to assess how effective different interventions might be for reducing exposure from food.

Some meat, dairy products and chicken eggs have been found to contain PFAS chemicals, but initial results from Food and Drug Administration research show that most detectable PFAS were found in seafood. Federal regulators recommend people avoid eating fish from waterways that have been impacted by PFAS. Produce grown with contaminated water can also become a source of exposure.

Take an inventory of products in your life that may contain PFAS  

Researchers are still studying how much consumer products influence the levels of PFAS in our bodies, and the National Academies report says there's not enough evidence to suggest that changing our use of consumer products would decrease our exposure.

But knowing which products contain PFAS can still help if you're trying to limit those chemicals in your life.

PFAS chemicals have been used widely in consumer products, in part because they are very good at making things waterproof, slippery or resistant to the elements. If you want to start looking for PFAS in your home, look for items that have those properties.

In the kitchen, you might find PFAS in nonstick frying pans, waffle irons, rice cookers or muffin tins. In your closet, they may be in waterproof coats or hiking pants. Carpets, upholstery, tablecloths and other household items treated to be durable, waterproof or stain-resistant could also contain PFAS.

Personal care products like dental floss, diapers, menstrual products, moisturizer and makeup are also potential sources of PFAS exposure, though some companies are discontinuing the use of those chemicals in their products.

Consider replacing items you use most

Dartmouth's Megan Romano says she usually recommends starting with the products that have the most contact with your skin.

"Look for moisturizers or foundations or bronzers, things that you're going to put on your face or your body and leave there all day," she says.

The use of PFAS in cosmetics may be declining, according to a FDA report from 2025. That report said the safety of most of the top PFAS chemicals used in cosmetics could not be determined, because the federal government does not have enough data on their toxicity.

If you're looking for PFAS-free products, try to figure out if a company has conducted third-party testing, meaning they've had their products tested by an independent lab.

When buying new products, you can look for the ingredient "PTFE" or other ingredients with "fluoro" in the name to figure out if PFAS chemicals have been added intentionally. The Consumer Product Safety Commission may be able to answer questions about PFAS in household goods.

These days, more and more manufacturers are phasing out PFAS. Chemical companies like 3M and BASF are stopping their production and use of the chemicals, and major brands like Dick's Sporting Goods and Sephora say they're working to remove or reduce PFAS in the products they sell.

The website PFAS Central has a directory of PFAS-free products, which includes more than 150 brands that say their products are PFAS-free.

And when major manufacturers cut down on PFAS use, that affects our health, too. Research shows that when the production and use of certain PFAS chemicals declined, the levels of those chemicals in the blood of Americans dropped over time. 

This story was edited by Malaka Gharib. The visual editor is CJ Riculan. We'd love to hear from you. Email us at LifeKit@npr.org. Listen to Life Kit on Apple Podcasts and Spotify, or sign up for our newsletter.

Copyright 2026 NPR

Florida is accusing OpenAI and its CEO Sam Altman of putting profit over safety, in the first lawsuit brought by a state against the ChatGPT maker over the alleged shortcomings of the chatbot.

The lawsuit, filed in Florida state court on Monday, claims the company and Altman failed to warn users that ChatGPT could be dangerous and instead marketed it as safe and reliable, including for children. It's the latest salvo in a growing effort across the country to hold artificial intelligence companies accountable when harms follow users' interactions with chatbots.

The lawsuit accuses OpenAI of aiding and abetting mass shooters, including a shooter at Florida State University who allegedly used ChatGPT to plan his attack, encouraging vulnerable people to commit suicide, and addicting children "to a tool that feigns human compassion to collect their data with no parental oversight."

"This litany of harms is driven by Defendants' insatiable quest to win the AI arms race and amass large fortunes, despite knowing the danger of ChatGPT," the complaint said. "The rise of OpenAI is attributable to a web of deceit and the exploitation of users (including Floridians), leveraging their data and safety to boost OpenAI's market value at unacceptable costs."

Florida is also seeking to hold Altman personally liable.

"Sam Altman and ChatGPT have chosen the AI race over the safety and security of our kids," Florida Attorney General James Uthmeier said at a press conference on Monday. He added he believes that Altman and the company could be liable "for potentially up to billions of dollars" in penalties.

"Losing a child is the most devastating tragedy that can happen to a family and we know that no words can come close to addressing the pain of such a loss," OpenAI spokesperson Kayla Wood said in an emailed statement to NPR.

"AI is a new and powerful technology, and we believe minors need significant protection, which is why we have put in place industry leading protections and policies," the statement continued. "In particular we built safety for minors directly into our products, including a more protective experience specifically for minors, an age prediction tool, defaulting users whose age we are not confident into our more protective experience, and giving parents tools to monitor their kids' use of AI."

The lawsuit alleges OpenAI's safeguards, including its parental controls, are inadequate and that the company has created "a dangerous public nuisance."

The first page of the lawsuit begins with a screenshot from OpenAI's website saying ChatGPT was "built with safety in mind." The image is followed by a footnote reading: "Not so."

Uthmeier's office is separately conducting a criminal investigation into OpenAI over the FSU shooter's alleged consultation of ChatGPT ahead of the April 2025 attack.

More than 20 lawsuits have been filed against OpenAI over harms allegedly stemming from ChatGPT use, including by families of victims killed and injured in a mass shooting at a school in Tumbler Ridge, Canada, in February, the family of a victim killed in the FSU shooting, and the families of seven people, including one teenager, who died by suicide or suffered delusions after using the chatbot.

Altman apologized to the Tumbler Ridge community in April, and OpenAI said in response to those lawsuits that it has a "zero tolerance" policy for using its tools to assist in committing violence.

After Uthmeier announced his investigation into the FSU shooting, an OpenAI spokesperson said that the chatbot "provided factual responses to questions with information that could be found broadly across public sources on the internet, and it did not encourage or promote illegal or harmful activity." The statement continued: "We work continuously to strengthen our safeguards to detect harmful intent, limit misuse, and respond appropriately when safety risks arise."

OpenAI has called the lawsuits over suicides and delusions "an incredibly heartbreaking situation" and said that it's working with mental health experts to improve how ChatGPT responds to signs of mental or emotional distress.

Other AI companies are also under legal scrutiny over how their chatbots have allegedly caused harm.

In response to a wrongful death lawsuit over the suicide of a Florida man who became attached to Google's Gemini chatbot, the company said: "Gemini is designed to not encourage real-world violence or suggest self-harm. Our models generally perform well in these types of challenging conversations and we devote significant resources to this, but unfortunately they're not perfect." It added that Gemini had "referred the individual to a crisis hotline many times."

In January, Character.AI settled multiple lawsuits brought by families who claimed its companion chatbots contributed to suicides and mental health crises among children and teenagers. The company said it "has taken innovative and decisive steps with regard to AI safety and teens, and will continue to champion these efforts and push others across the industry to adopt similar safety standards." That includes barring users under 18 from interacting with or creating chatbots.

Last month, the state of Pennsylvania sued Character.AI, alleging its chatbots posed as doctors and offered medical advice, in violation of state medical licensing rules. A Character.AI spokesperson told NPR at the time the company doesn't comment on pending litigation, but that its "highest priority is the safety and well-being of our users."

Copyright 2026 NPR

Stay up to date on daily news with NPR's Up First newsletter sent every weekday morning.

Anthropic, maker of Claude, one of the most popular artificial intelligence models on the market, has started the ball rolling on an initial public offering.

The company announced on Monday that it has "confidentially submitted" a draft registration statement known as a Form S-1 with the Securities and Exchange Commission for its proposed IPO.

"This gives us the option to go public after the SEC completes its review," it said in an announcement. "The proposed initial public offering will depend on market conditions and other factors."

Anthropic said the number of shares and price have not yet been set.

The filing sets the stage for what is shaping up to be a trio of potential mega-listings by tech companies: Elon Musk's SpaceX, involved in rocketry and AI, filed financial information in late May for an IPO. And OpenAI, the maker of ChatGPT, is also widely reported to be planning a listing.

"We believe this represents an opening of the floodgates for the IPO market, which has been relatively dormant for a few years, with these three major conglomerates set to go public later this year, but this has turned into a race to reach public markets over the coming months," analysts at Wedbush Securities wrote in a note after Anthropic's announcement.

Anthropic's is already valued at almost $1 trillion

The filing comes a week after Anthropic announced that its valuation had reached $965 billion after its latest round of fundraising. That makes it bigger than OpenAI, which in March said it was worth $852 billion after a recent funding round.

It also comes at a time of great hype and anxiety over AI. Shares in chip stocks and other companies involved in the AI supply chain have been soaring, as companies like Anthropic and OpenAI have been investing heavily in data centers.

But recent polls show that Americans are wary of AI's impact on daily life.

Copyright 2026 NPR

WASHINGTON — A novel pill helped people with advanced pancreatic cancer live longer, researchers reported Sunday, raising hopes of long-needed better treatments for one of the deadliest types of cancer.

"While not curing the cancer, it is a very large step forward," said Dr. Zev Wainberg, of the University of California, Los Angeles, who helped lead the study.

The drug is called daraxonrasib and it blocks a mutated protein that fuels tumor growth in more than 90% of pancreatic cancer cases — a target that had eluded treatment for decades.

The daily pills nearly doubled survival time, with fewer severe side effects, in a study that randomly assigned the experimental drug or more chemotherapy to 500 patients whose metastatic, or spreading, cancer had quit responding to prior treatment. The findings were published in the New England Journal of Medicine and presented Sunday at the American Society for Clinical Oncology meeting in Chicago.

Those taking daraxonrasib lived for a median of 13.2 months compared with 6.7 months for chemotherapy recipients. While that may seem like a small improvement, Wainberg said it marked the first drug to show a substantial advantage over chemotherapy.

"Having treated pancreatic cancer for 16 years, I actually started crying" when first seeing the study results, Dr. Rachna Shroff of the University of Arizona Cancer Center, who wasn't involved with the research, said from the ASCO meeting. She was struck by how "patients stayed on this treatment because it was providing durable and meaningful benefit to them."

The pills' effects eventually wane but recipients used them for significantly longer than the comparison group stayed on chemotherapy, reporting less pain and a better quality of life as their tumors shrank. Many still were using the drug after the data was analyzed, which Wainberg said means the survival gap may widen as researchers continue tracking them.

Dr. Brian Wolpin, of the Dana-Farber Cancer Institute, presented the findings Sunday. He said the drug should become "a new standard of care" for previously treated metastatic pancreatic cancer, adding that researchers also will explore its use earlier in the disease, including to see if tumor shrinkage might let more patients qualify for surgery.

Side effects most likely to affect pill usage were a rash that can be severe and mouth sores, he said.

Maker Revolution Medicines funded the study and the Food and Drug Administration plans to expedite review of the drug. Meanwhile, the agency is allowing what's called "expanded access" to the experimental drug for patients who meet certain criteria. The drug garnered public attention when former U.S. Sen. Ben Sasse described on "60 Minutes" how he's had less pain while taking it. Oncologists are being flooded with requests as the special access program gets started.

Pancreatic cancer is among the most deadly forms in large part because it's hard to detect before it starts spreading to other organs. The American Cancer Society estimates about 67,000 new cases will be diagnosed in the U.S. this year and more than 52,000 people will die from the disease. The five-year overall survival rate is 13%.

Unlike with other cancers that have benefitted from a variety of chemotherapy alternatives, pancreatic cancer has been harder to tackle.

Cancer specialists not involved in the new research expressed optimism that this may be a turning point in the quest for new options, with dozens of experimental drugs in development.

The new drug targets mutations in the RAS gene family that normally regulates cell growth. So-called KRAS mutations are especially critical in fueling pancreatic cancer. But a structure that made it hard for drugs to stick to the mutated proteins meant this cancer driver was long considered "undruggable."

Revolution Medicines' drug uses what's essentially a molecular glue to bind with multiple KRAS subtypes. Wainberg said researchers next will probe whether the drug worked better in certain of those subtypes.

The drug will change pancreatic cancer treatment, said Dr. Andrew Coveler of the Fred Hutchinson Cancer Center, who wasn't involved in the research.

"This thing works drastically differently," he said.

Wainberg said other drugs in development target specific KRAS subtypes. Other approaches in earlier stages of testing include vaccines designed to prevent recurrence after pancreatic cancer surgery by teaching the immune system to recognize the mutated protein.

Copyright 2026 NPR

How do you make explosives using household items? How do you make meth? How do you plan a school shooting? If you ask the popular AI chatbots most people are familiar with, chances are they will say that it's illegal, harmful or that answering would be a policy violation.

But another type of AI model will never refuse to provide what the user asks for. In recent months, these models have become more accessible and popular.

"Everybody can download and operate their own state-of-the-art model and use it for great things and terrible things," said Noam Schwartz, CEO of Alice, an AI security company that has conducted red-teaming and safety evaluation for AI model developers.

Teaching models when to say "no"

Big AI companies such as OpenAI, Google, Anthropic and xAI train their proprietary models to refuse requests deemed as harmful or inappropriate. Legions of workers instruct models when and how to refuse certain prompts.

These methods don't always work and carry pitfalls: some harmful requests go through, while other users complain about innocuous requests being refused. Chatbots that initially say "no" can be manipulated into saying "yes" using cleverly phrased prompts, such as posing them as poems. Even with guardrails, popular chatbots have been used to plan mass violence and generate deepfake child sexual abuse material. In some instances, parents have accused AI chatbots of encouraging their children to harm themselves.

But there's a whole other class of AI models whose guardrails are much easier to strip away. They're known as open-weight models. Some are made by tech giants, such as OpenAI and Alibaba, while others are put out by smaller outfits like China's DeepSeek. Like their better-known proprietary counterparts, many possess advanced capabilities such as writing functional code or generating life-like images. Unlike with ChatGPT, Claude or Gemini, it's easier to permanently remove their built-in safety guardrails – and the companies behind them have no idea how they're being used.

Getting rid of open-weight models' guardrails used to take time and deep expertise. But in recent months, that process has become dramatically more accessible and popular.

Recent method makes removing model guardrails easier than ever

Safety guardrails of open-weight models can be weakened or removed in many ways. This is largely because the model developers have made what's known as the model weights available to the public. Model weights are sets of parameters, like knobs and dials in a machine, telling the models how to process information.

One recently developed method called "abliteration" has caught the attention of AI and national security researchers. By tweaking model weights, people can take away the model's ability to say "no."

Hugging Face, which hosts open-source AI models, currently lists over 6,000 abliterated models, compared to about 600 in 2024. On Hugging Face, abliterated models currently outnumber models that have their guardrails removed using other methods, according to research by the National Counterterrorism Innovation, Technology, and Education Center (NCITE), a Department of Homeland Security-supported research consortium based at University of Nebraska at Omaha.

What's more, new tools are making it much easier to create abliterated models. "That was [the job of] the data scientist, you know, a senior employee" at a leading AI lab, said Schwartz. "Now, everybody with access to the internet and a laptop for like 400 bucks can actually run this thing on their own machine."

One such tool is Heretic, which automates the abliteration process. All a user has to do to remove a model's guardrails is to give Heretic two lines of instructions, and the process can take as little as a few minutes. The application has gotten more popular on the code repository GitHub since February, according to Alice's research.

Some lawmakers are taking notice. In late April, House lawmakers attended a demonstration of abliterated models hosted by NCITE, Politico reported.

"[What] was frightening about this demonstration was how readily available some of this content or software is on kind of the black market right now, and how it can be weaponized and used to manipulate people, destroy lives and build weapons of mass destruction," said Rep. Andy Ogles (R-TN) in a video put out by Republicans on the House Homeland Security Committee. Hugging Face and GitHub, where abliterated models and the tools to make them can be downloaded, are not black markets.

Models without guardrails can be both useful and dangerous

It is difficult to get a comprehensive picture of how people are using open-weight models, because these models are run locally on users' computers, and don't need the internet to function. Unlike with proprietary models, the model developers cannot monitor what users are asking the models.

But there's growing anecdotal evidence for how people are experimenting with altered models.

Several accounts on X said they have used abliterated models to generate pornography.

An individual in a pro-ISIS chat room claimed they used an "uncensored" AI to research the amount and type of explosives needed to destroy "Trump Tower in the U.S.," according to the Counter Extremism Project, a nonprofit that focuses on counterterrorism.

On one cybercrime forum, a user asked for ideas to get around an AI model's guardrails so they could use AI to make scam calls. Another user recommended Heretic, according to research by Alice.

While giving users information on how to conduct harmful activities could be concerning, the more worrying part is how the chatbots can egg users on, said Samuel Hunter, senior scientist and director of academic research at NCITE.

"It's jarring when you see it in real time, this sort of bubbly persona with some of the abliterated models that's like, 'Oh, what a great idea to create this bomb,'" Hunter said. "Imagine somebody that has no other kind of social connection and it starts to take them down a darker path and really encourage them."

There are legitimate uses for AI models without guardrails, such as using them to catch bad actors and to help with cybersecurity research, said Schwartz, the AI security company CEO. Law enforcement may use a modified model to simulate possible terrorist attacks, said Hunter.

Philipp Emanuel Weidmann, the developer of Heretic, said AI is just an information processing and retrieval system akin to a search engine, which can be used in many ways. The fact that criminals use them is "a corollary of what AI models are: namely, tools," he told NPR.

When it comes to safety guardrails, "there's this very small set of entities that decide what is acceptable and is not acceptable," Weidmann said, referring to the big AI companies making proprietary models. "That creates a stifling intellectual climate that I do not want to work in."

For now, open-weight models are not as capable as the most advanced closed-weight models. But their capabilities are less than one year behind, according to the recent International AI Safety Report commissioned by the British government and led by computer scientist Yoshua Bengio.

The capability gap may matter in areas like cybersecurity, where the most advanced closed-weight models, such as Anthropic's Mythos and OpenAI's GPT-5.5, are starting to get good at not only spotting vulnerabilities, but also writing code to exploit those vulnerabilities. In the arms race of cyber offense and defense, companies using closed-weight models to screen and patch vulnerabilities may still have a leg up compared to attackers using open-weight models, security researchers say.

Mitigating the risks from models without guardrails comes with tradeoffs 

One line of mitigation focuses on making guardrails more tamper-proof. Early research shows that filtering out content related to making biological weapons from AI training data can reduce how often the model responds with information that could be used for harm.

Another line of mitigation focuses on restricting access to models without guardrails. Model-hosting platforms like Hugging Face can limit access to models specifically trained for "harmful purposes," according to the International AI Safety Report.

The same report also recommended that model developers evaluate their models' potential for harm prior to release.

These measures come with flaws and tradeoffs, according to the report. "Features enabling beneficial applications in medicine or research can be repurposed for harm, and once weights are public, distinguishing legitimate from malicious uses can be difficult," it says.

Weidmann, the creator of Heretic, is working to make sure his tool can remain accessible to the public in the event that platforms like Hugging Face take down abliterated models.

"There's too much power in AI," he said. "Unrestricted models being available to the powerful while not being available to anyone else will lock in power structures forever."

Copyright 2026 NPR

KINSHASA, Congo — The head of the World Health Organization arrived in Congo's capital, Kinshasa, late Thursday to witness efforts against an outbreak of a rare type of Ebola virus, as medical personnel struggle with a lack of equipment, a distrustful population and armed groups in a volatile region.

"To come here is to really show to the community that they're not alone," WHO Director-General Tedros Adhanom Ghebreyesus told reporters at the airport.

"Pushing orders from my comfortable office in Geneva is easy, but I'm asking my colleagues to work with the community and I am asking communities to protect themselves," he added.

Medical aid donated by the European Union arrived in Ituri province, the heart of Congo's Ebola outbreak, on Thursday. The United States announced the same day $80 million in additional aid, bringing its total commitment to more than $112 million.

Health workers with scant supplies have been struggling to contain an outbreak of the Bundibugyo virus, a kind of Ebola that has no approved treatment or vaccine. In some areas, doctors have resorted to wearing expired medical masks while treating suspected patients.

According to WHO, 1,077 suspected cases and 238 suspected deaths have been recorded as of Tuesday,

Dangers faced by health workers have been heightened by anger among residents over the stringent medical protocols for dealing with the bodies of victims, which clash with local burial rites. Residents have launched at least three attacks against health centers.

Tedros said other challenges are also complicating the containment of the outbreak, including the high number of people displaced by armed conflict in the region, and food insecurity.

On Wednesday, he had called for a ceasefire in a region where armed groups have staged violent attacks for decades.

"We cannot build community trust or isolate the sick while bombs are falling," Tedros said.

Tucked in the northeastern part of Congo close to the Ugandan border, Ituri province has been reeling from attacks by the Allied Democratic Force, a rebel group allied with the Islamic State group, and a coalition of ethnic militias. In early May, the ADF killed at least 40 people and burned several homes in Ituri.

The illness also has been reported in the Congolese provinces of North Kivu and South Kivu, south of Ituri, where the Rwanda-backed M23 rebel group controls many key cities, including Goma and Bukavu. The rebels have reported two cases.

The region's main airport in Goma, which doubles as a staging ground for humanitarian efforts into the region, has been closed since January 2025, when M23 seized the city.

The conflict has precipitated one of the world's largest humanitarian crises, with at least 7 million people displaced in eastern Congo.

The WHO chief said Thursday he discourages countries from imposing travel bans against nationals of countries affected by the outbreak.

"There are ways to manage workers and to manage cases without having a strong, restricted travel ban and we don't encourage that as WHO," Tedros said.

The Trump administration last week announced a temporary ban on the entry of people without U.S. passports, as well as U.S. green-card holders, who have visited Congo, Uganda or South Sudan in the past 21 days. It said Wednesday it plans to send Americans who are exposed to Ebola to a new facility in Kenya instead of flying them to the U.S. Congo's neighbors, Uganda and Rwanda, recently closed their borders.

Copyright 2026 NPR

CAPE CANAVERAL, Fla. — A rocket belonging to Jeff Bezos' Blue Origin exploded during a test at the launch pad Thursday night, shaking nearby homes and briefly painting the sky orange.

Blue Origin said its New Glenn rocket exploded during an engine-firing test being conducted ahead of a satellite launch planned for next week. No one was hurt, according to officials at Cape Canaveral Space Force Station.

"It's too early to know the root cause but we're already working to find it," Bezos said via X. "Very rough day, but we'll rebuild whatever needs rebuilding and get back to flying. It's worth it."

The massive New Glenn was grounded in April after it left a satellite in the wrong orbit because of engine failure. It was only the third flight of the rocket that Blue Origin intends to use to launch landers to the moon for NASA, including the landers that will take astronauts to the lunar surface.

The company had been on track to launch a prototype lunar lander to the moon on a flight test this fall. Earlier this week, the space agency awarded Blue Origin a contract worth hundreds of millions of dollars to launch a pair of moon buggies in the next few years as part of the Artemis program.

"Spaceflight is unforgiving, and developing new heavy-lift launch capability is extraordinarily difficult," NASA Administrator Jared Isaacman said via X. He promised to provide information on any impacts to the Artemis program, including the moon base that he recently outlined.

Homes shook in nearby Cape Canaveral and Cocoa Beach around 9 p.m., with residents turning to social media to wonder what happened. Launch Complex 36 is visible from the beach, and the internet quickly filled with photos of an orange fireball.

The rocket was supposed to blast off next week with internet satellites that are part of the Amazon Leo constellation in orbit.

Emergency crews remained more than an hour after the explosion. Officials stressed there was no threat due to fumes or other potential hazards.

Space Force officials said the explosion would not affect upcoming launches by other companies from other pads. United Launch Alliance's Atlas V rocket is due to blast off Friday night with a batch of Amazon Leo satellites, the same kind that this rocket was supposed to take up.

SpaceX's Elon Musk, who's had his own share of rocket explosions, offered his condolences. "Sorry to see this, I hope you recover quickly," he told Blue Origin via X.

Towering at 321 feet (98 meters), New Glenn made its debut in 2025. It is named after John Glenn, the first American to orbit Earth, and is much bigger and more powerful than the New Shepard rockets that have carried tourists to the fringes of space from Texas.

Copyright 2026 NPR

California's attorney general sued the genetic testing company formerly known as 23andMe on Thursday, alleging it failed to protect sensitive user data in a 2023 breach that affected nearly 7 million people across the country.

Attorney General Rob Bonta filed the lawsuit against Chrome Holding Co., which 23andMe rebranded under after filing for bankruptcy last March. 23andMe is known for its direct-to-consumer DNA test kits that provide customers with information on their ancestry and genetic predispositions for certain health conditions.

The lawsuit calls for various civil penalties against 23andMe and injunctions blocking the company from further violations of California’s privacy protection laws.

The company has acknowledged that it suffered a major security breach in 2023 that resulted in about 14,000 accounts being accessed, through which they were able to steal the data of nearly 7 million customers. The cyberattack utilized “credential stuffing," which takes advantage of customers' tendency to use weak or common passwords or reuse passwords between multiple accounts.

Bonta's office said this was a well-known attack that businesses should know to guard against. The attackers used stolen user account credentials, including ones from a massive data breach in October 2017 that affected MyHeritage, one of 23andMe’s former partners. After that breach, 23andMe did not take common protocols such as asking customers to reset their passwords or use multifactor authentication.

23andMe did not immediately respond to an emailed request for comment.

“23andMe’s security measures were so lax that the threat actor was able to operate undetected within 23andMe’s systems for over five months, and remarkably, 23andMe only began investigating after the threat actor offered the stolen user data for sale on the dark web and reached out to 23andMe to demand a ransom,” prosecutors said in the complaint.

In October 2023, the stolen data appeared for sale on the dark web, with the poster specifically touting that about 1.1 million consumers’ data belonged to Asian-Pacific Islander and Ashkenazi Jewish users.

“The sale of this data on the dark web took place amidst a period of mounting anti-Asian American and Pacific Islander and antisemitic hate and violence,” Bonta said in a press release. “This is disturbing and incredibly dangerous.”

Some of the data stolen included raw genetic data, health reports, DNA shared with other relatives, and locations and birth years of relatives.

The lawsuit says that after notifying the public about the breach, 23andMe continued to mislead consumers about the severity of the breach and the company's role in it.

The company has said it only found out about the breach in October 2023 when the stolen data was posted for sale on the dark web. However, the lawsuit said the company failed to properly investigate red flags that appeared months earlier, such as a “suspicious spike in user login attempts” in July and a Reddit post discussing a possible breach and sale of user data in August.

Genetic data requires “one of the highest levels of protection,” and California law “mandates a heightened legal obligation” to protect it, the lawsuit said.

Bonta also intervened to ensure customers' genetic data wouldn't be mishandled during 23andMe's Chapter 11 bankruptcy and asset sale, arguing that California's Genetic Information Privacy Act required companies to obtain opt-in consent from customers before selling their genetic information to third parties. However, the sale was allowed to proceed.

In 2024, 23andMe agreed to pay a $30 million settlement in a class-action lawsuit accusing the company of failing to protect customers whose personal information was exposed in the breach. The amount was raised to $50 million to resolve most U.S. customer claims and received final approval in January by a federal judge overseeing 23andMe's bankruptcy.

For a decade, UC San Diego students have been running Triton Gaming Expo, serving up a multiverse of fun that includes an artist alley, cosplay, industry and voice-actor panels, and a promise of insane tech.

"Triton Gaming Expo is our annual gaming convention," said Jewelle Tatad, the external vice president for Triton Gaming. "We're basically taking over all the open space in the Price Center. We're going to be having fundraisers downstairs as well as some partnerships with the bookstore. We're going to be having the majority of our programming on the second floor, with the East Ballroom housing our different gaming clubs as well as some of our partner organizations and programs such as esports. The West Ballroom will be having some of our artists, and up on the fourth floor we will be having our cosplay café."

Magnolia Hyon is the executive officer for ArtSpark Pre-Professional Art Club.

"ArtSpark is a pre-professional art club, the very first at UCSD, and we're focused on helping students break into the art industry," Hyson said. "So becoming professional artists, whether that be in animation or video games, we're just trying to help them get art education, kind of more structured education than you get here in the art program, and build their portfolio through many of our portfolio projects, and build community. So as a pre-professional art club, being around art in the industry is really important. So we're going to be tabling with some of our members' work. We can share what we've done as a club and what our individual members have done in the company of these really huge game industry giants like Riot, Supergiant, Behemoth. It's really valuable for our members to get networking with people like that."

And gaming company professionals welcome the chance to be on campus.

"Some of them are also UCSD alum — and some of them are even Triton Gaming alum— a lot of them are really excited to see what kind of art talent, as well as marketing talent, as well as all this other stuff that is applicable to this industry, is on display," Tatad said. "A lot of people in Triton Gaming do want to go into the games industry, whether that's on the esports side or in the video game side, where they are creating games. So I think that the talent is all super duper excited to be there just to talk to these people and kind of see like where students are at."

Hyon points to the work of yet another student organization, the Video Game Development Club.

"The whole club's base comes together to work on one game," Hyon said. "I was a huge part of that this quarter. It was an incredible experience, really helped me build my portfolio and make connections with the members. The Video Game Development Club is just constantly running projects and game jams to really be that video game presence on campus where curriculum kind of falls short."

UC San Diego students are quick to organize and take action when they see a need. Since the university is best known as a STEM school, Tatad wants to highlight its more artistic side.

"As someone who runs a student organization that is so focused around gaming, which I think is in and of itself an art form, I think it's really important to kind of uplift these student artists, especially in a school where I feel like we kind of get sidelined," she said.

But artists won’t be sidelined this weekend. In fact, attendees will be rewarded for visiting artist booths.

"A lot of artists organized this thing called Stamp Rally," Tatad said. "This one is specifically transportation theme centered. We were making a Riso-printed map specifically for it. So that'll be organized by some of our artists. If you make a purchase from each of these artists, you get that stamp as a special prize."

Tatad also wants to make attending a convention affordable for students.

"I feel especially in an era where like con culture and fandom culture is becoming more and more gatekept and more and more inaccessible, I think providing these spaces for people to be able to engage in gaming, engage in anime, engage in art is super duper important, especially for university students, which many of us are low-income," she said. "Many of us don't have this access to fandom without having to shell out a lot for it."

Hyon agreed: "I think it's really helping or supporting con culture here. People love art and media, and it helps bring people closer together through their interests and even through their specialty careers like voice acting and video game development, art, of course, and just being able to put it all towards this big event and then having so many people involved with it. Then getting the engagement, and engagement supports the community. I think it's a message to students — not necessarily from the university, but from students — you're not alone and we're all here together and we all love this stuff and we can bond over that."

Triton Gaming Expo takes place Saturday and Sunday at UC San Diego's Price Center. The event is free for UCSD students and $12 for the public for a two-day pass.

NEW YORK — Happy, a Bronx Zoo elephant who gave researchers new insight into the animal's behavior and became the crux of a closely watched animal rights case, has been euthanized at age 55, the zoo said Wednesday.

The Asian elephant was put to sleep Tuesday at the zoo where she lived for almost a half-century. Zoo officials said some age-related conditions accelerated in recent weeks, and she showed signs of a falloff in kidney or liver function. A necropsy revealed arthritis and large, inoperable uterine tumors that are impossible to diagnose in elephants through exams or imaging, the zoo said.

"She was a wonderful elephant," interim zoo director Craig Piper said in an interview Wednesday, as heavy-hearted staffers absorbed the loss of an animal some had tended for over 30 years. "She served as a tremendous ambassador for elephants and for elephant conservation."

Since Happy's death, the zoo's 57-year-old Patty is the last elephant on exhibit in the United States' largest city. The zoo's parent institution, the Wildlife Conservation Society, decided 20 years ago to stop acquiring pachyderms.

Born in the wild in Asia, Happy was brought to the U.S. as a 1-year-old. She was named for a character from "Snow White and the Seven Dwarfs" before arriving at the zoo in 1977.

Happy keenly engaged with her keepers and was easy to motivate with favorite treats, such as watermelon or strawberries, said Keith Lovett, the zoo's director of animal programs. Piper said she sometimes stashed treats in her ear to save for later.

In 2005, she showed researchers that elephants can recognize themselves in a mirror — a sign of self-awareness seen in only a few other species. During the experiment, Happy faced her reflection and repeatedly used her trunk to touch an "X" painted above her eye, a mark she could see only in the mirror.

She was paired with other elephants until her last partner died in 2006. Happy then lived separately from Patty and a third elephant out of concern that they wouldn't get along, though Lovett said the animals could see, smell and touch each other over a divider. The third elephant, called Maxine, died in 2018.

Zoo officials said the median life expectancy for Asian elephants in U.S. zoos is about 45 years. Their life expectancy in the wild is more difficult to pinpoint.

During Happy's lifetime, zoo elephant exhibits came under increased scrutiny. Some experts said urban animal parks were too small for creatures that roam extensive distances in the wild. Animal rights activists argued that zoo enclosures were no place for big-brained, social pachyderms.

Some zoos phased out their exhibits and sent elephants to sanctuaries, though some other zoos remain committed to keeping and breeding the creatures, arguing that they help keep people interested in saving wildlife.

One activist group, the Nonhuman Rights Project, sued the Bronx Zoo in 2018, seeking to have Happy declared a "person" for legal purposes and moved to a large animal sanctuary. It was the first such case about an elephant, according to the group.

Citing a principle that's used to challenge the legality of a person's imprisonment, the activist group said Happy was "an extraordinarily cognitively complex and autonomous nonhuman being" who was unlawfully deprived of her liberty and suffered from being pent up in a exhibit without other elephants.

Zoo officials said Happy was assiduously cared for and had space for swimming, foraging and other natural behavior. Uprooting her from her longtime home could harm her, the zoo said.

New York's top court ultimately rejected the activists' claim, by a 5-2 majority. Colorado's highest court later issued a similar ruling about five elephants in a zoo there.

Still, two of the New York high court judges wrote pointed dissents. One called Happy's captivity "inherently unjust and inhumane" and "an affront to a civilized society."

The Nonhuman Rights Project has continued pursuing cases about elephants in various other states.

Happy spent her final weeks, by her choice, in an off-exhibit barn and yard within her enclosure, Piper said. In a zoo version of hospice care, staffers provided hydration, nutrition and pain management, he said.

Meanwhile, Patty is doing well, the zoo said.

The Wildlife Conservation Society said in 2006 that once there was only one elephant, the animal might be moved to another zoo if circumstances were right. Piper said the zoo will be "really thoughtful and careful" in contemplating whether to move Patty from her home of 53 years.

Copyright 2026 NPR

Updated May 27, 2026 at 6:15 PM PDT

A Google software engineer has been charged with using confidential company information to make $1.2 million on Polymarket, in the second known federal criminal case connected to lucrative trades on a prediction market site.

Michele Spagnuolo, 36, an Italian citizen who lives in Switzerland, was arrested on Wednesday and charged with commodities fraud, wire fraud, money laundering and other counts for allegedly placing bets on search trends based on internal Google data that tracked user searches.

"Unlike the counterparties to his trades, Spagnuolo knew the outcome of these wagers before the trading public did because he had accessed Google's confidential, commercially valuable internal data," according to the federal indictment, which authorities unsealed on Wednesday.

Prosecutors say Spagnuolo, operating under the username AlphaRaccoon, placed a number of wagers on Google's most-searched person of 2025.

He bet nearly $1 million that Kanye West's wife, Bianca Censori, would not be the most-Googled person. He also bet more than $600,000 that Pope Leo XIV would not take the top spot, and he placed another wager that the rapper D4vd would be the No. 1 most-Googled individual at a time when most Polymarket traders "assigned near-zero probability" to the singer, who has been charged with murder.

In all, according to prosecutors, Spagnuolo bet $2.7 million on 25 separate outcomes in the Google search market, netting $1.2 million in profit.

The charging documents say once Spagnuolo transferred his winnings out of his cryptocurrency wallet, he removed the name AlphaRaccoon from his Polymarket account.

The Commodity Futures Trading Commission brought a separate civil case against Spagnuolo for allegedly violating commodities law.

Spagnuolo did not return a request for comment.

Google said in a statement that the company cooperated in the federal government's investigation into Spagnuolo, who has been placed on leave.

"The employee accessed our marketing material using a tool available to all employees, but using such confidential information to place bets is a serious breach of our policies," said Google spokesperson Jaclyn Vazquez.

Olivia Chalos, Polymarket's chief legal officer, said in a statement that it "is the only prediction platform to date whose cooperation has led to insider trading charges in the United States," adding that, since users on the site use crypto to trade, it is "transparent, traceable and bad actors leave footprints."

While the laws that apply to the prediction market industry are less strict than stock market rules, what's commonly understood as "insider trading," or abusing non-public confidential information for profit, is illegal under federal law.

But in prediction market forums on messaging sites such as Discord, users scour markets for large, unusual trades and encourage others to follow those bets with their own wagers.

"AlphaRaccoon has alpha," said one user on Discord, using the slang term for any information that gives you an edge on prediction markets, pointing to Spagnuolo's large bets on the most-Googled person of the year before Google had released it. "Check AlphaRaccoon account" said another user when asked how they should bet on the market.

Prediction markets sites such as Kalshi and Polymarket have erupted in popularity in President Trump's second term, allowing anyone to bet on company announcements, geopolitical events, the outcome of art auctions, elections and a seemingly endless array of other topics.

And as more and more people seek profits in every facet of modern life, online sleuths have increasingly identified wagers that appear too good to be true, suspiciously confident long-shot bets that have netted prediction market traders six- or seven-figure profits.

In one such instance last month, a master sergeant with the U.S. Army Special Forces was charged with using classified information about the capture of Venezuelan leader Nicolás Maduro to rake in more than $400,000 on Polymarket.

The Spagnuolo indictment was unsealed a day after President Trump vowed on Truth Social to allow the prediction market industry to "thrive" by asserting federal regulators' "exclusive authority" over the controversial betting sites.

For months, administration officials have been fighting state officials in court over who should police the prediction market industry.

State officials say the platforms are essentially gambling operations and should be subject to state gambling rules, whereas the Trump administration views Polymarket and Kalshi as offering a type of "futures contract" that falls under the umbrella of the Commodity Futures Trading Commission, which has historically overseen markets on things like grain futures, crude oil and precious metals.

Polymarket's most popular platform, which is based in Panama, is technically inaccessible to American users. It was forced to shut down its U.S. operation in 2022 as part of a settlement with federal regulators who said the site was operating without a trading exchange license.

Two years later, the FBI raided the apartment of the company's founder, Shayne Coplan, as part of a probe into whether Polymarket was violating that agreement.

The Trump administration dropped that investigation. Trump officials even invited Coplan to the White House for a summit on cryptocurrency.

The president's oldest son, Donald Trump Jr., is an advisor to Polymarket and Kalshi, and a partner in 1789 Capital, which is a major investor in Polymarket.

Copyright 2026 NPR

NEAR TALLAHASSEE, Fla. — Nicole Dahrouge is not a salamander. Crouched in a bog, hands searching through short grass, she states it aloud for the second time in the last hour, almost like an affirmation.

"I mean, I'm not a salamander," she says. "But if I was, I would lay eggs right there."

There's always a bit of urgency collecting frosted flatwoods salamander eggs. The tiny and secretive ground-dwelling salamander is one of the most imperiled amphibians in North America, teetering on the brink of what biologists call an "extinction vortex" — the point at which a plant or animal's population is so small that its problems start to fatally compound.

Dahrouge's job at the Amphibian and Reptile Conservancy (ARC) is to keep the "frosties," as they're lovingly called, from slipping over that precipice to the point of no return; to bolster their population and buy them time to adapt to the fast-changing world.

That starts with duck-walking through clingy brush to find their eggs.

"It's just like the world's itchiest scavenger hunt interspersed with little periodic injections of serotonin when you find something fun," she says, rubber boots squelching in damp earth.

Frosted flatwoods salamander eggs require a very specific set of climatic conditions to hatch. They're laid, each fall, in ephemeral ponds; on dry mounds, like the one Dahrouge is circling, that should be inundated by winter's rains. It's an annual gamble for the salamanders. Without inundation, the eggs will dry up. And with weather patterns shifting, as the world warms, it's a wager they're less and less likely to win.

Dahrouge is trying to skew the odds. Left in the wild, the survival rate of eggs is very low. Once they hatch into aquatic larvae, their chances don't get much better. Drought aside, "everything eats them," she says. "They're just like little protein gummy bears."

So Dahrouge is trying to help get as many of them through their first phases of life as possible by raising them in captivity. It's a monumental effort. A "stop-gap," she says, with no end in sight. And it's an example of how difficult it is to recover a federally threatened or endangered species when they are already on the brink.

"Man, when we let species get to this point, it's so much effort and so much work and so much resources to get it to a point where it's back," says JJ Apodaca, ARC's executive director.

"But we either do this now or we watch them go extinct."

Few places left to live

Most of what scientists know about frosted flatwoods salamanders comes from what they're able to glean during egg-laying season. As a part of the mole salamander family, they spend most of their lives in burrows underground.

Fittingly, their color ranges from deep black to dark chocolate, their backs covered in crisscrossed and mottled white-gray lines. "Think like a dewy sparkly spiderweb laid over a black background," Dahrouge says. "They're beautiful."

In 1999, the U.S. Fish and Wildlife Service (FWS) listed the flatwoods salamander as threatened under the Endangered Species Act. A decade later, realizing that there were actually two species of "flatwoods salamanders" that looked very similar — the reticulated flatwoods salamander and the frosted flatwoods salamander — they were split up under federal law. The reticulateds were listed as endangered, the frosties as threatened.

A status review of the frosted flatwoods salamander, published by the FWS in 2019, found that they too warranted being listed as endangered due to "declining population trends." Seven years later, the FWS still hasn't taken any action.

"Even though it hasn't officially been reclassified yet, it still gets nearly all the same protections as an endangered species under the Endangered Species Act," a spokesperson for the FWS said in an email.

Advocates for the salamander say it should be uplisted to reflect the reality, noting that the Trump administration has sought to limit certain protections for threatened species.

On the ground, ecologists like Dahrouge and Apodaca are more focused on stabilizing populations and improving habitat than the salamanders' federal status.

"Policy can't go out and save a species," Apodaca says. "We, as a community, we as a society have to go out and save that species."

For both flatwoods salamanders, that means improving habitat. They live in the longleaf pine forests of the Southeastern U.S. — flat, open stands of wildfire-dependent trees, grasses and shrubs that used to cover the coastal plains from southern Virginia to east Texas.

Then came large-scale logging. Agriculture. Subdivisions. A century of fire suppression.

Today, just 3% or so of longleaf pine forest remains intact in the Southeast — made up of scattered patches that the frosties, reticulateds and other imperiled longleaf-loving animals are left clinging to.

"It's a globally imperiled ecosystem sitting on top of a global biodiversity hotspot," says Houston Chandler, science director for the nonprofit Orianne Society, which focuses on amphibian and reptile conservation. "Not a great combination."

Chandler has been working in the largest remaining stand of old-growth longleaf pine forest, on Eglin Air Force Base, to improve habitat for the reticulated flatwoods salamander by mechanically removing undergrowth and restoring wetlands.

It's brutal labor, Chandler says, often done in the heat of summer. And it requires constant maintenance. But, he says, it's also working. They have more sites occupied by reticulated flatwoods salamanders at the military installation than ever before.

But they're still vulnerable.

"It took decades and decades of fire suppression and poor habitat management and land conversion for them to become endangered in the first place," Chandler says. "So it's not going to be an overnight fix."

Raising "beefy" salamanders

Boots squelching, Dahrouge shifts to point at a cluster of frosted flatwoods salamander eggs she's found. It's a glistening glob of goo.

"My perhaps not entirely appropriate description when I'm training new people is it looks like someone hocked a loogie into the base of this plant," she says.

The eggs are maybe a couple of weeks old. The salamanders are still squiggling commas inside them. With a pocketknife and tablespoon, she loosens up the patch of dirt they sit on and scoops it from the ground. In severe drought years, like the kind the Southeast is experiencing now, they'll pull every egg cluster they can find — "a salvage," as they call it. To prevent the eggs from drying up, they're stored in plastic containers packed with damp earth and taken, miles away, to a place that's more climate-controlled.

That place just happens to be sitting in her backyard.

"This started as a woodworking shed that I built for my projects," Dahrouge says, walking through double doors into a garage-sized shack. She gestures toward a pile of unused saws, lumber and recreational equipment that's been pushed into a corner. "Now we have salamanders."

The eggs are stored until they're ready to be inundated with water and primed to hatch. Dahrouge checks on them constantly. "I'm a helicopter parent, 100 percent," she says.

The next phase requires moving them to a series of cattle tanks that line her backyard. There are dozens of them. Each is its own miniature artificial wetland — a mesocosm — built with vegetation and water that Dahrouge and her colleague, Matthew Goetz, collect from the field.

They also collect food. At least once a week, Dahrouge or Goetz goes to nearby wetlands, stirs shin-deep water and collects silty samples filled with arthropods, daphnia and other macroscopic critters the salamanders will eat. Hours more are spent picking out pinprick-size predators from the water with pipettes.

"It's very time-intensive but very necessary," Dahrouge says. "I don't know if you've ever seen a dragonfly larva, but they look like the creature from Alien and can unhinge their jaw and eat a salamander larva that's the same size as they are, so we want to save them from that fate."

The end goal is to raise as many healthy larvae and salamanders as they can, so they can be released back into the wetlands they were found in or to bolster others. In conservation circles, the technique is known as headstarting — raise an animal in captivity, release it in the wild.

The practice was used to pull the California condor back from the brink of extinction and is used for many imperiled species. It can be "hugely powerful to keep a cohort alive," says Carola Haas, an ecologist at Virginia Tech who's worked with reticulated flatwoods salamanders.

"But any time you rear something in a tank, you're selecting for captivity. And the characteristics that make you good at surviving in a tank may be the exact opposite of the characteristics that make you good at surviving in the wild," she says.

Instead, she argues, the conservation community should focus its efforts on restoring the salamanders' habitat.

"If the habitat restoration doesn't happen, nothing can persist," she says. "And that's expensive and time-consuming enough in itself."

Currently, frosted flatwoods salamanders are only known to exist in four areas. Some populations are isolated, making them vulnerable to inbreeding. All are at risk of stand-alone weather events like hurricanes, disease or drought.

With so little habitat and so few frosties left, Apodaca says, they've crossed the point where habitat preservation and restoration, alone, will not be enough to keep them from going extinct.

"In my opinion, there's zero chance this species makes it out and naturally recovers itself if we just fix the habitat," he says.

The same, he says, is true for many other imperiled amphibians and reptiles.

"We have to be, by necessity, entering into a new era of conservation that I think of as the age of implementation," he says. "There's been decades of arguments of how active we should be in intervening [in nature] … to get over the next step, we've got to do a lot more direct species interaction."

At the back corner of her erstwhile woodworking shed, Dahrouge lifts a small, glistening female frosted flatwoods salamander from a bed of moss. She normally lives in a "bog garden" that Dahrouge and Goetz built in the backyard, over the course of months, where the hope is to eventually facilitate captive breeding.

All of the captive adults have been named after stars, because the white flecks on their black stomachs look like constellations in the night sky. Andromeda, Dahrouge says, is one of her favorites. Not that she has favorites, she quickly adds.

"Ninety-nine percent of the people in the world will never see this animal," she says. "And I wish that everyone could because they're just so infinitely worth knowing."

Copyright 2026 NPR

CAPE CANAVERAL, Fla. — NASA is already ordering landers, rovers and drones for a sprawling moon base, less than two months after the Artemis II's record-breaking lunar flyaround.

The space agency outlined the first phase of its moon base plans on Tuesday, awarding hundreds of millions of dollars in contracts to four U.S. companies.

Jeff Bezos' Blue Origin will provide a pair of landers to deliver moon buggies to the lunar surface, at a spot near the moon's south pole. These so-called lunar terrain vehicles will be built by Astrolab and Lunar Outpost. Firefly Aerospace, which landed successfully on the moon last year, will deliver the first drones to the moon.

All this hardware is ideally supposed to arrive before the first Artemis astronauts land on the moon, planned for as early as 2028.

During April's Artemis II mission, four astronauts flew around the moon, traveling deeper into space than the Apollo moon crews did during the late 1960s and early 1970s. For next year's Artemis III, another team of astronauts will practice docking NASA's Orion capsule in orbit around Earth with the lunar landers being developed for crews by Blue Origin and Elon Musk's SpaceX.

NASA is targeting Artemis III for mid-2027, with a landing by two astronauts following as soon as 2028. The moon base's second phase, from 2029 into the early 2030s, will start building up the permanent infrastructure, including a power grid. As for when the base will be ready to support astronauts for extended periods in specialized permanent habitats, that's expected sometime in the 2030s, during the third phase.

"Then we'll be able to say, 'Hey, we're permanently here and we're not giving it up,'" said NASA's moon base program executive Carlos Garcia-Galan.

Garcia-Galan envisions a moon base sprawling over hundreds of square miles, with a perimeter marked by drones, dubbed MoonFall, stationed at the corners.

NASA Administrator Jared Isaacman said these territory markers are meant to be respectful of other countries' spacecraft and equipment that might be nearby. He expects reciprocity in the matter.

The goal of the moon base is to encourage a lunar economy while conducting scientific research and laying the foundation for a Mars expedition, Isaacman stressed.

"For those waiting patiently, the grand return is close at hand and we will not slow down," Isaacman said. "We are really just getting started."

Copyright 2026 NPR

“I'm an avid cyclist, so I ride through here all the time,” said Leaf. “And as I was coming down Bonita Road, I just saw it. And that's actually what I call this pattern, is the ‘it’ pattern. I just saw it. And I knew that was the answer.”

The “it” pattern is a series of panels painted a glossy blue and purple, with little shapes cut into the panels and pulled outward. It was installed in April. Now, thanks to the vivid color, the museum is hard to miss.

Leaf said part of his goal was to have the outside of the museum reflect what’s going on inside.

“This is how we let not only the community know, but the world know, ‘Hey, there's a museum over here,'” he said.

What’s on the inside is a history of not only Bonita, but the Sweetwater Valley in general.

“People don’t know about Bonita, maybe about South Bay, so they’ll get a little history about the Sweetwater Valley, a little bit about the first people here, the Kumeyaay, as well as moving into the early days of water with the Sweetwater Dam, and then going into the agriculture, some of the citrus that was in the region,” said museum executive director Wendy Wilson.

The museum has just two rooms. In one, you’ll find the history. The other is where you find the cultural side of things. Various exhibitions come and go through the year, including works by Michael Leaf.

Opening on June 6 is "Beyond the Border, Plein Air.” It features artists from around the region who paint outdoor scenes across the county, capturing the bright light, coastal vistas and urban scenery in real-time.

In Leaf’s Chula Vista studio, a few miles away, you'll find various works of art wrought out of metal in various stages of completion. Much of it will eventually end up in the Bonita Museum.

One piece that will eventually be on public display is what you might call Leaf’s magnum opus. It is so large it has its own room in his studio.

“This is the second to-scale Last Supper on planet Earth. I just felt like making it, and when I learned the scale and size of the painting, I couldn’t believe the numbers. It was 29 feet by 15 feet,” Leaf said. “I thought I misread the paper because I’ve not seen the painting in real life.”

Leaf said at this point, he can’t reveal where the piece will end up.

Back at the museum, the metal artist is finishing up work on what will be the first piece in what Wendy Wilson said will become a “grand outdoor sculpture garden.”

Leaf calls it the piece Inner Strength. It’s a metal figure of a human holding two diverging steel bars in place. Leaf said it’s a metaphor of just how strong we as a people are, both externally and internally; a beacon of resilience.

You could also say an example of a homegrown native of the South Bay sharing his artistic gift to make his corner of our region a more beautiful, inviting place.

Leaders of the California State University system, the CSU, want it to become the nation's first artificial intelligence-powered institution of its kind.

It entered into a $17 million no-bid contract with OpenAI last year to provide students, faculty and staff with a new resource: ChatGPT Edu — a version of the popular generative AI chatbot intended for use by educational institutions. The system recently renewed that contract for another $13 million a year for the next three years.

"No other university system in the U.S. or internationally is doing anything like this, not at this scale," said Mildred García, the CSU's chancellor, during a February 2025 press conference announcing the partnership.

But in a recent survey, majorities of its students and faculty said they were skeptical of the benefits of AI for education, and they worry about AI's impacts on job security, creativity and the environment.

Colleges across the country – from Syracuse University to Dartmouth College to the University of Minnesota – have inked similar deals with AI companies, but as the largest public four-year system in the U.S., the CSU's partnership stands out.

As higher ed scrambles to figure out the benefits and harms of AI, the CSU offers an early look at what happens when an administration commits to a technology that its own community isn't convinced will improve education.

What the university has to gain

In December 2024, university leaders flagged a potential partnership with OpenAI as "a huge branding opp[ortunity]," according to an internal CSU planning document obtained by NPR.

Ed Clark, chief information officer for the CSU's office of the chancellor, told NPR in an email that "the planning document demonstrates the extent to which the CSU thoughtfully approached selecting a vendor that could support our commitment to innovation, accessibility and academic excellence."

Clark said the system chose to partner with OpenAI because they offered "the most cost-effective option that could make it even possible to bring AI tools to more than a half a million students, faculty and staff."

A separate document obtained by NPR, dated 2025, shows that the CSU expected questions around its partnership with OpenAI. The document, titled "Potential follow-up questions on ChatGPT Initiative," advises officials to explain the no-bid contract by saying the deal is "essential for the success of the CSU's AI strategy."

The document goes on to say, "After conducting extensive research and evaluating various AI tools and vendors, it was determined that OpenAI is uniquely positioned to meet our needs."

AI won't be used to teach classes, and Clark, the CSU spokesperson, told NPR in an email that the technology should supplement learning, not replace it. Both the CSU and OpenAI frame AI adoption as a necessity to prepare students for careers steeped in this technology.

"As they prepare for the workforce, AI literacy is becoming part of career readiness… so the CSU's role is to help students understand how AI is changing their disciplines and how to use it ethically and responsibly," Clark said.

Leah Belsky, vice president of education at OpenAI, told NPR they share a responsibility to "help students use these tools well… to harness their full potential and succeed in the AI-driven future of work."

But Martha Kenney, a professor and science and technology scholar at San Francisco State University, part of the CSU, says some faculty and students reject the idea that AI in higher ed is an inevitability, and that their perspective deserves consideration.

"I think refusing this technology needs to be a position that's on the table," Kenney says. She says rejecting this technology on campuses is justified, given generative AI's environmental impact and the use of copyrighted work to train models. She also questions the educational value of technology like ChatGPT Edu: She says offering a chatbot that allows students to take shortcuts on assignments is "cheating our students out of an education."

Kenney co-authored a petition that called on the CSU not to renew its contract for ChatGPT Edu.

But Clark says the "online petition does not reflect overall sentiment from within our community." He says the CSU's survey shows strong support for AI given that majorities of students and faculty say it has had a positive impact on their learning and work.

Clark also says the CSU chose to renew its agreement with OpenAI after its generative AI advisory committee, which is composed of students, faculty and staff, "unanimously recommended renewing the contract."

How students, staff and faculty feel about AI

The CSU serves about 470,000 students and, according to the system, it awards nearly half of all bachelor's degrees in California. Its student body is diverse: Roughly half are Hispanic, more than a quarter of undergraduates are the first in their family to attend college and many students work while they attend school.

Last fall, the system invited students, staff and faculty across all 22 campuses to take a survey on their views around AI. More than 94,000 people responded, and the results show widespread use of generative AI tools, but also significant ambivalence about the technology.

The survey did not ask whether students, faculty and staff agreed with the system's decision to spend millions on a contract with OpenAI.

Among the survey's topline findings:

• More than half of students and around 6-in-10 faculty and staff report using AI regularly for coursework and tasks related to their jobs.
• Roughly 65% of students and 59% of faculty said they were skeptical AI was benefitting education overall.
• 80% of students said they wouldn't be comfortable turning in AI-generated work as their own. 
• About 64% of students said AI has "positively affected" their learning, while about 35% said AI "negatively affected" their learning. 
• About 56% of faculty reported AI had positively affected their teaching, research and administrative experience. But in a separate survey question, 52% reported a negative effect.

Roughly 84% of students said they used ChatGPT. About a quarter of them said they used the version provided by CSU and the vast majority said they used the free version.

Large majorities of students and faculty also worry about AI's impact on creativity (83% of students, 82% of faculty), job security (82% of students, 78% of faculty) and the environment (80% of students, 84% of faculty).

The survey had some limitations, says David Goldberg, an associate professor at San Diego State University, part of the CSU, and one of the survey authors.

"The findings are based on the people who did respond. We don't know the opinions of the people who didn't," Goldberg explains. Still, he says the responses are a "pretty good representation across different fields of study and across different demographics."

Goldberg says the survey illustrates a tremendous amount of nuance in opinion across all groups.

"Even within one student, you can be using the tool a lot, see real advantages, and at the same time see these negatives," he says.

What students stand to gain – and lose – from AI on campus

Sejal Daterao is one of those students with complicated feelings.

Daterao, 30, says she enrolled in the information systems master's program at California State University, Long Beach, to learn how to use AI more efficiently.

As a student, she says she uses ChatGPT Edu and other AI tools to conduct research, summarize text and video lectures, and create quizzes targeted to the subjects she's studying.

And she says she's grateful the CSU provides access to ChatGPT Edu — which includes features not available on the free version of ChatGPT. As a grad student, she says it would be hard for her to foot the bill for a premium subscription.

"Helping students use such technologies firsthand is really a good thing, honestly," Daterao says.

But she doesn't describe herself as pro-AI.

She's frustrated by the occasional false information AI chatbots generate and by tech companies' use of creative work to train AI models without providing credit and compensation to artists.

"It has a lot of bad sides, and a lot of good sides," Daterao says. "If you are smart, if you are being ethical, you can use the good sides in a really amazing way."

But another student, H, doesn't see many redeeming qualities to AI. She's in her fourth-year studying computer science at San José State University, also part of the CSU, and she asked that NPR refer to her by only her first initial because she's actively applying to tech jobs and doesn't want her opinions on AI to impact her employment prospects.

H says she was annoyed when she noticed that her classmates were using AI to write assignments for them.

"It was pissing me off, which is why I completely avoided using it at first," she says.

Eventually, H says she began using AI chatbots for "menial tasks" like writing emails, and then to help her on coding assignments.

But she noticed that when she used AI to code, "I found that I was using it more as a crutch instead of actually helping. So that was one of the telltale signs that I should stop using it.'

She says her resistance to AI has only deepened as she's learned about the environmental impacts of data centers.

H understands that the CSU is under pressure to adapt to an emerging technology. But she says she's "a little disappointed that they accepted it with open arms immediately."

And H worries that pushing AI use in coursework will prevent students from learning the foundational skills they need to be successful.

"It's something I've struggled with," she says. "Trying to use it to learn basics kind of led to just not learning basics, but using it to avoid putting in effort."

Educators "can't ignore the technology"

Zach Justus, a communications professor and director of faculty development at California State University, Chico, part of the CSU, has spent the last few years encouraging faculty to adapt their teaching to the AI age — which means experimenting with the technology to figure out what it can and can't do.

He says he's excited about the innovative ways some faculty members are leveraging and allowing students to use AI. But he says adaptation, in certain circumstances, also includes redesigning coursework to prevent AI-use.

"The most important thing that we tell faculty is that they cannot ignore the technology," Justus says. "If we ignore it, we are not doing our jobs."

He says he understands the critiques of the university system's contract with OpenAI – including the argument that the system shouldn't spend millions on an AI chatbot when it's facing budget cuts. But he says it's also a problem if only some students can afford the premium versions of this software.

Without the system providing these tools to students, Justus says, "You're just systematically advantaging students with more financial resources, and that's crappy."

English professor Jennifer Trainor isn't ignoring AI, but she's not a fan of it either.

Trainor, who teaches at San Francisco State University, says her approach is to teach students about AI and the ethical questions it raises. She says she safeguards the learning process from AI by requiring students to brainstorm and draft by hand during class time. And she allows students to use AI to edit their writing, but she requires them to reflect critically on the changes it made.

"I am really trying to get them to do their own writing and thinking," Trainor says. "And I'm also giving them chances to see what happens when they use tools to improve their writing and thinking."

Some students refuse to engage with AI altogether, Trainor says. She describes it as a "groundswelling of resistance" on the campus.

"They're ethically opposed to the environmental impacts and the bias and the erasure of their jobs and voices and creativity. [They] don't like it," Trainor says.

But it's clear that, for now, it's not going anywhere.

This reporting was supported by a grant from the Tarbell Center for AI Journalism and the Omidyar Network's Reporters in Residence program.

Edited by: Nicole Cohen

Copyright 2026 NPR

The first time Elizabeth Smart stepped on stage at a bodybuilding competition, she was terrified.

She says her smile froze. Her hands shook. Every movement had been choreographed and practiced over and over again, down to the turns and poses she would hit beneath the bright stage lights.

But there was only so much she could do to prepare for the pageantry. Unlike in training, she was wearing oversized costume jewelry, including a large ring. The blonde hair extensions were new, too.

Then, as she flipped her hair over her shoulder, the ring snagged one of the extensions.

"I just ended up ripping through the extension and just taking out a chunk of my hair, and then turning around and smiling," she says, laughing about it now.

At the time, she says, she wanted to run offstage.

Instead, she kept posing in towering heels as the judges rated the body she'd spent years trying to survive inside.

For Smart, bodybuilding isn't about the trophies. Yet, four competitions and several medals in, she's earned something she never expected: confidence in her body.

"I'm at a point in my life where I want to celebrate it," Smart says, "I don't want to carry shame about my body."

A traumatic detour

In 2002, Smart was just 14 years old when a self-proclaimed prophet abducted her at knifepoint from her Salt Lake City bedroom while she slept beside her younger sister.

For months, the world watched the search for her unfold. Her face was plastered across television screens and the front pages of newspapers. All the while, she was living in the woods just miles from her home.

Now, at 38, Smart remembers the ways she tried to survive the nine months she was held captive and repeatedly sexually assaulted. She endured frequent humiliation and psychological manipulation.

In her latest book, Detours, Smart describes trauma as a detour — a path you never planned for and never wanted. She's says she survived captivity in part by holding onto small memories and moments that reminded her that her life existed outside those woods.

"My body was hurt, and it had felt like it had been crushed," she says. "But it carried me through."

Disconnecting from the body

That kind of positive relationship with the body after trauma can take years — and sometimes decades — for survivors to develop, says Robyn Brickel, a licensed therapist in Virginia who specializes in trauma-related disorders.

"When early childhood trauma happens, especially sexual trauma, people disconnect from their bodies because it's unsafe," Brickel says. "That's how they survive."

During the abuse, some victims mentally leave their bodies, focusing instead on small details in the room, she says.

"Lots of trauma survivors will tell you, 'I know exactly how many light bulbs there were in the chandelier,' how many cracks were in the ceiling, the pattern on the wallpaper" while the abuse was occurring, she says. "Because that's where they are."

She says the body becomes something to escape rather than inhabit. For many survivors, that disconnection doesn't disappear once the abuse ends.

Brickel says survivors often struggle with feeling shame, confusion and betrayal connected to the body.

"Lots of survivors believe their bodies betrayed them," she says.

Smart says she understands that feeling.

Raised in a conservative Mormon home, where modesty and purity were heavily emphasized, Smart says she struggled with profound shame after the abuse. She spent much of her time playing the harp, avoided boys and had few close friends.

For years, after she was back home, she says she felt pressure to become what she describes as "the most innocent of victims," she says. "I had to always do the right thing, always say the right thing."

By the time she was rescued in 2003, nine months after she was kidnapped, millions of people already knew her name and face. Unlike many survivors, Smart had to heal while in the public eye.

Today, Smart says, she sees herself differently.

"I can be an advocate for women and children," Smart says. "But I also can step on stage in a bikini and strut around and strike a pose. And that's OK."

To Brickel, that shift — from invisibility to visibility — is significant.

"Trauma survivors will [often] make themselves as unattractive as possible to not get attention," she says. "They want to disappear. Be invisible."

'There's no finish line'

Smart says her relationship with exercise has changed dramatically over the years.

After she was rescued, she says she occasionally ran but didn't stick with it. She eventually became a marathon runner, though recurring knee pain forced her to stop.

"I always need a goal and I need a deadline," she says.

Bodybuilding offered both. So, she started strength training about a year and a half ago.

Now she trains at least five days a week, for about 45 minutes at a time. She tracks her meals carefully, counts macros and walks roughly 10,000 steps a day, often on an incline treadmill.

Mounting research shows weight lifting may help some trauma survivors reconnect with their bodies in healthy ways. According to a study published last year in Frontiers in Psychology, resistance training was linked to reduced post-traumatic stress disorder symptoms and improved emotional well-being. And a 2023 study published in the same journal found that many trauma survivors described weight lifting as empowering — saying it helped them rebuild confidence, regain a sense of control and feel safer in their own bodies again.

Still, Brickel says physical training and trauma recovery don't always intersect in healthy ways. For some survivors, exercise becomes another form of disconnecting rather than healing — similar to how some use drugs, self-harm, eating disorders or overworking as a way to outrun emotional pain.

The difference, Brickel says, often comes down to intention and emotional awareness.

"Can I think and feel at the same time?" she says. "Am I running from something, or am I adding to my life?"

That question sits quietly beneath much of what Smart describes. She talks less about perfection than presence. Less about punishment than appreciation.

One of her favorite book passages comes from Charlotte Brontë's 1847 novel Jane Eyre. Smart describes Mr. Rochester telling Jane he could crush the cage around a bird, but never destroy the bird itself.

Smart says that metaphor stayed with her.

Though her body felt broken, she says, "it never let my soul be destroyed. It carried me through my kidnapping. It gave me three beautiful children."

Then she says something that still surprises her: "My body is incredible."

For Brickel, positive statements like that can represent years of emotional work. "We work on that in therapy all the time," she says.

But she also notes that healing is rarely linear. Some survivors speak about their trauma right away. Others wait decades. Some never talk about it at all.

"There's no finish line," Smart says. "I hope I never stop progressing."

These days, Smart says she's seriously considering another bodybuilding competition later this year in Nashville — an all-female event that recognizes women who have survived trauma.

Her face lights up as she talks about it.

Not because she believes trauma disappears, but because she no longer wants survival to be the only lens through which she sees herself.

"We can be lots of things," she says.

When she doesn't feel like walking outside during training season, Smart climbs onto her treadmill and watches The Great British Bake Off while dreaming of sweets.

"I want that," she says, laughing. "I am adding that to my post-show treat list."

"And I want the whole thing," she adds. "Not just a slice."

Copyright 2026 NPR

Updated May 28, 2026 at 12:41 PM PDT

The United States is heading into the height of disaster season, with millions of people — and billions of dollars worth of property — at risk this summer from hurricanes, wildfires and flash floods.

When homes burn or flood, insurance is supposed to soften the blow. But homeowners and renters often discover what their policies actually cover only after catastrophe strikes.

With some preparation, though, you can better your odds of a smoother recovery if extreme weather hits your home.

Here are three things you can do now to make sure you're ready to deal with insurance after a disaster.

How much coverage do you have? 

First, check how much your insurance would pay out if your home is destroyed.

Ideally, it's enough to cover the cost of rebuilding. However, underinsurance is a chronic problem in the U.S., according to United Policyholders, a national consumer advocate. The group says its surveys routinely show that half of homeowners don't have enough coverage to replace their homes after a disaster.

"So check, not just with your insurance company, but if you know somebody in the building industry that can tell you what it costs per square foot these days to rebuild a house, do the math and check if you have enough coverage, because a lot of people don't," says Douglas Heller, insurance director at the Consumer Federation of America.

You should also check your deductible. That's how much money you would have to pay out of pocket if your home is damaged.

"When you increase your deductible, you can get some savings" on home insurance, Heller says. "But you are transferring the risk of a devastating storm back into your bank account and away from the insurance company."

And verify that your homeowners or renters insurance covers living expenses if you're displaced.

"Think about how much it would cost if you had to be out of your house for three weeks, six weeks, three months," Heller says.

For renters in single-family homes, make sure your landlord's home insurance is up to date.

And remember, flooding isn't covered by homeowners and renters insurance, so you need a separate policy for that. Most people who buy flood coverage do so through the National Flood Insurance Program.

Get your documents in order

Next, document your valuables. Having an up-to-date record is a big help if you have to file an insurance claim. Heller suggests using a cellphone to record a video every year of the stuff inside your home or apartment.

Marcus Coleman, vice president of community resilience strategy at United Way, also recommends storing a list of emergency contacts and any medications you need on the cloud or a thumb drive.

Having easy access to copies of your insurance policy, identification and financial information like credit and debit cards is a good idea, as well, according to the Federal Deposit Insurance Corp., an independent government agency.

"If I just have a split second, I know where to go to get what I need to be able to answer some of the questions that might come up over the next days or weeks after a disaster," Coleman says.

Keep up with property maintenance

The third thing you need to do is maintenance work around your property, such as trimming trees and clearing roofs and gutters of flammable materials, like leaves and branches.

Insurance policies don't typically state that homeowners have to do things like cut back tree limbs that hang over their house, says David Boohaker, a lawyer who represents policyholders in disputes with insurance companies. "But what it will say is, 'We don't [have] to pay you for damages due to neglect or your failure to care for your home,'" he says.

And insurers have found ways to monitor houses they insure, often without homeowners knowing. "One thing that has come up very frequently, and often it surprises consumers and they get angry about it, is the extent to which insurers are using drones now to do flyovers of property, which means they are looking for these things now," says Brendan Bridgeland, director of the Center for Insurance Research, a national consumer advocate.

In addition to routine maintenance, insurance experts point to programs in a number of states that are designed to help homeowners protect their homes from disasters. In Alabama, for example, homeowners can get insurance discounts if they install roofs that are designed to withstand high winds.

"If there is bad weather, then you [may not] have to make a claim in the first place if you have a higher building standard or higher protection for your residence," Bridgeland says.

Homeowners can contact their state's insurance department to find out if there are programs to help pay for upgrades to make houses less vulnerable to extreme weather.

And if you need help with things like disaster recovery and housing assistance, you can call 211 for information about resources in your area.

Copyright 2026 NPR

Some 50,000 residents of Garden Grove, California remain under an evacuation order Sunday as emergency response teams struggle to deal with a potentially explosive situation at a nearby aerospace manufacturing plant.

Here's the latest on what's happening at the plant, and what could yet come.

Overnight, the chemical tank appeared to have cracked

The tank, which is located in the southeastern corner of the GKN Aerospace facility in Garden Grove, California, holds somewhere around 7,000 gallons of methyl methacrylate, a highly toxic, highly flammable chemical used in the manufacturing of resins and plastics.

Late Saturday, firefighters approached the tank to "get eyes" on what was happening, TJ McGovern, interim county fire chief for the Orange County Fire Authority, said in a post on social media. "What they found was a potential crack in the tank."

A cracked tank could actually be good news, sort of

The incident at GKN Aerospace began Thursday, according to emergency responders. It remains unclear exactly what went wrong, but the chemicals in the tank began to exceed a safe temperature.

Methyl methacrylate is a clear, colorless liquid that is highly volatile and releases energy exothermically when it reacts. If that reaction occurs in a container, then it can cause a sudden pressure build up, effectively turning the container into an explosive.

That's exactly what happened at a plant in the United Kingdom in October 2009. According to one report, an explosion at a resin manufacturing site involving the chemical destroyed the factory and caused blast damage to adjacent buildings. Nobody was killed, but windows were blown out as far as 600 ft. from the blast site.

The tank in California suffered damage and had already begun to bulge outward, according to Craig Covey, the incident commander at Orange County Fire. The tank is being continuously sprayed with water to keep it cool. It sits next to two other tanks, one of which was safely drained and neutralized, and the other of which seems stable for now.

Speaking on Friday, Covey told reporters that if the damaged tank cracked, it could actually be the best outcome. While far from ideal, it would be better if the 7,000 gallons leaked out slowly rather than sparking an explosion inside the tank.

"In a weird world that's the best case scenario believe it or not," he said. "Because once it comes out it is no longer an explosive hazard."

There are still many environmental and health hazards

The company which owns the plant, GKN Aerospace, said on Sunday that it was "working around the clock to mitigate the risk of a leak."

The risks are numerous – methyl methacrylate can irritate the skin and eyes. It can also cause respiratory issues and neurological symptoms including headache and lethargy. More long-term exposure can cause lung and organ damage.

There's less information available on what the environmental impacts might be should the chemical spill, but they're unlikely to be good.

In anticipation of a possible leak, "we've created containment barriers," Covey said on Friday. Firefighters have already laid down sand and other materials to try and stem the flow of the chemical into nearby storm drains and waterways, he told reporters.

For now, residents must remain evacuated

"We know that you're out of your homes, we want to get you back, but we cannot do that until it's deemed safe," Orange County Fire Chief McGovern said in Sunday's video statement.

Residents are being evacuated to protect against both the potential of a large blast and fumes that an explosion would release. Because the fire department can't predict which way those fumes might travel, they're having to evacuate a large area around the plant.

So far, no fumes have been detected by firefighters or the Environmental Protection Agency, which has set up monitoring stations around the site.

McGovern said the reconnaissance gave him reason to hope that the newfound crack "could potentially be relieving some of the pressure in there."

"We're not there yet, but this was a step in the right direction," he said of the recent assessment of the tank. "And there could be a lot more coming shortly."

Copyright 2026 NPR

Think the Olympics — with a multi-million jackpot and no rigid doping rules.

The Enhanced Games are happening on Sunday in Las Vegas with the promise to create a global sports competition "where elite athletes push the limits of human performance," according to organizers.

The privately-funded games have garnered considerable attention online, and are financially backed by Donald Trump Jr.'s venture capital firm, 1789 Capitol, and tech billionaire Peter Thiel. The company behind the event, the publicly-traded Enhanced Group, is using the games as a launch pad for their business, selling peptides and other supplements. Enhanced is also documenting the effects of the drugs on athletes for its own research.

"I understand that there's a very large commercial opportunity for this company, but it is something I think borders on the lines of ethics," said Dr. Aaron Baggish, a professor of medicine at the University of Lausanne in Switzerland who worked with sports teams in Boston for 20 years.

The one-night competition will be held in custom-built arena at Resorts World Las Vegas with a four-lane 50-meter pool, a six-lane sprint track and a weightlifting stage.

What will happen at the Enhanced Games?

Over 40 athletes from across the globe will compete across three sports: swimming, track and weightlifting. There will also be a "strongman" deadlift showdown..

The four swimming events are 50-meter and 100-meter races in both freestyle and butterfly. The track and field races include the 100-meter sprint and the 100-meter and 110-meter hurdles. The weightlifting series will feature the clean and jerk as well as the snatch events.

The total prize pool for the games is $25 million, CEO Max Martin told Front Office Sports News. Each individual event has a total purse of $500,000, with a $250,000 prize awarded to first place. The company also said that if an athlete breaks a world record in either the 100-meter sprint or the 50-meter freestyle, they will be given an additional $1 million. Though, neither mark would be recognized by governing bodies, such as World Athletics, which requires drug testing.

Enhanced Games competitors are allowed to use performance-enhancing drugs that are banned in internationally recognized sporting events. According to the company, the drugs are FDA-approved and prescribed by doctors.

Enhanced did not break down what specific athletes used which drugs, but they announced on Wednesday in the lead-up to the event that 91% of the athletes competing used testosterone or testosterone esters, 79% used human growth hormone, and 62% used stimulants, such as adderall.

Baggish said the public should understand that "FDA approval does not equate with safe use when the drug is not used the way the FDA has approved it."

He added that there is a difference between someone being prescribed testosterone because they don't produce enough versus someone taking much larger doses for athletic achievement.

Research on high doses of drugs like testosterone, for example, have shown an increased risk of heart disease, Baggish said. He said the number of Enhanced athletes using testosterone is "very concerning."

"We have to be careful not to confuse short-term success with long term implications," Baggish said. "These athletes, I assume many of them, if not all of them, will go through the Enhanced Games without any visible problems whatsoever."

"But what happens to them three years from now, five years from now?" Baggish said.

For context, the International Olympic Committee follows strict guidelines on doping with the World Anti-Doping Agency list of prohibited substances. These include testosterone, human growth hormones, and stimulants such as adderall. There are a few strict exemptions.

Who's behind the games?

Enhanced Games was founded by Aron D'Souza, an Australian entrepreneur and lawyer who previously worked with Thiel to file litigation against Gawker Media. Martin, the CEO of Enhancement, took over for D'Souza in November.

In 2023, D'Souza announced the Enhanced Games, and last May at a launch event he divulged the details. At the event, he said the idea for the games came to him in 2022 "during a moment of reflection."

"I imagined a new kind of competition, one where science and sport and society could evolve together, where we stopped apologizing for progress and started to embrace it," D'Souza said.

His company is exploring athletic enhancement "openly, responsibly and ethically," he added.

The administration of the drugs was overseen by independent medical and scientific commissions, he said. D'Souza said the drugs would be administered with "fully personalized protocols and constant monitoring to ensure their safety and peak performance."

Baggish, the professor of medicine, said he was approached to join Enhanced Games as a doctor which he said he immediately turned down. He said the apparent medical oversight should not blind people to the dangers of performance-enhancement drugs.

"That's akin to me saying: If I'm a physician watching you smoke cigarettes, I can make smoking safe for you," Baggish said.

Representatives for Enhanced Games did not immediately respond to a request for comment on the criticism of the games. In a 2024 statement responding to the International Federation of Sports Medicine's concerns over the games, D'Souza said the competition's approach "aims to reduce the risks associated with unregulated practices and bring them under medical oversight."

"The Enhanced Games do not promote the indiscriminate use of restricted substances," D'Souza said. "Instead, we advocate for the safe, responsible and clinically supervised use of performance enhancements."

Which athletes are participating? 

There are some big names competing at the games.

U.S. Olympic gold-medalist swimmers Cody Miller and Hunter Armstrong, a former world-record holder, along with British Olympic silver-medalist swimmer Ben Proud will compete. U.S. runner Fred Kerley, a three-time world champion and two-time Olympic medalist, will be a sprinter at the games. Kerley was banned from competing in globally recognized competition for two years after missing drug tests — a violation of anti-doping rules that does not necessarily mean an athlete is taking drugs — the Athletics Integrity Unity announced in March.

Armstrong and Kerley have said they are among the athletes who won't be using performance-enhancing drugs in the games.

Most of the participating athletes trained for the competition in Abu Dhabi as part of Enhanced's own study.

Former NFL player Emmanuel Acho and MLB Network broadcaster Abby Labar will be the anchors for the competition.

What do the experts say? 

The games have been largely panned by outside medical experts and sports governing bodies. Multiple recent studies assess the harm surrounding the Enhanced Games.

Travis Tygart, the CEO of the U.S. Anti-Doping Agency, called the games a "dangerous clown show that puts profit over principle" in a statement. The International Olympic Committee said the games are a "betrayal of everything that we stand for." The World Anti-Doping Agency (WADA) last year urged U.S. authorities to stop the games.

The International Federation of Sports Medicine said in 2024 that they see the medical oversight as "insufficient" to support the athletes.

"There's simply no way to make the use of these drugs safe by medical monitoring in the short and long term," Baggish, the professor of medicine, said. "It's essentially a natural history experiment to see what happens."

Dr. Michael Joyner, a clinical anesthesiologist whose research focuses on physiology and human performance, said that there's a credible body of research showing the effects of doping across a wide spectrum of people, including a study linking the premature deaths of bodybuilders to the use of steroids, and archival data from East Germany in the 1960s.

"The question is: What is anybody trying to prove?" Joyner said. "It's been known for years that anabolic steroids work," he added, "I don't understand what's new here."

Copyright 2026 NPR

SpaceX launched its biggest, most powerful Starship yet on a test flight Friday, an upgraded version that NASA is counting on to land astronauts on the moon.

The redesigned mega rocket made its debut two days after SpaceX CEO Elon Musk announced he's taking the company public. It blasted off from the southern tip of Texas, carrying 20 mock Starlink satellites that were released midway through the hourlong spaceflight that stretched halfway around the world.

The spacecraft reached its final destination — the Indian Ocean — despite some engine trouble, before erupting in flames upon impact. That last part was not unexpected, according to SpaceX.

Musk called it "an epic" launch and landing.

"You scored a goal for humanity," he told his team via X.

It's the 12th test flight of the rocket that Musk is building to get people to Mars one day. But first comes the moon and NASA's Artemis program.

NASA Administrator Jared Isaacman flew in for the launch, saying Starship is now one step closer to the moon.

The last of the old space-skimming Starships lifted off in October. SpaceX's third-generation Starship — a souped-up version dubbed V3 — soared from a brand-new launch pad at Starbase, near the Mexican border. Last-minute pad issues thwarted Thursday evening's launch attempt.

SpaceX was hoping to avoid the fireworks it experienced during back-to-back launches last year when midair explosions rained wreckage down on the Atlantic. Earlier flights also ended in flames.

There was no fireball this time until the very end. The spacecraft plummeted upright into the Indian Ocean under seemingly full control, then toppled over and ignited.

While the liftoff itself went well, not all of the engines fired as the booster attempted a controlled return. The spacecraft also had to make do with fewer engines, but kept heading eastward 120 miles (194 kilometers) up. A pair of modified, camera-equipped Starlinks ejected from Starship provided brief views of the spacecraft in flight — a remarkable first.

At 407 feet (124 meters), the latest model eclipses the older Starship lines by several feet (more than 1 meter) and packs more engine thrust.

The revamped booster sports fewer but bigger and stronger grid fins for steering it back to Earth following liftoff, and a larger and more robust fuel transfer line to feed the 33 main engines. This fuel line is the size of SpaceX's Falcon 9 first-stage booster. The retro-looking, stainless steel spacecraft also has more of everything — more cameras and more navigation and computer power — as well as docking cones for future rendezvous and moon missions.

Starship is meant to be fully reusable, with giant mechanical arms at the launch pads to catch the returning rocket stages. But on this latest trial run, nothing was being recovered. The Gulf of Mexico marked the end of the road for the redesigned first-stage booster, and the Indian Ocean for the spacecraft and its satellite demos.

NASA is paying SpaceX billions of dollars — and also Jeff Bezos' Blue Origin — to provide the lunar landers that will be used to land Artemis astronauts on the moon.

The two companies are scrambling to be first.

While Starship has reached the fringes of space on multiple flights lasting an hour at most, Bezos' Blue Moon has yet to lift off, although a prototype is being readied for a moonshot later this year.

NASA is following April's successful lunar flyaround by four astronauts with a docking trial run in orbit around Earth planned for next year. For that Artemis III mission, astronauts will practice docking their Orion capsule with Starship, Blue Moon or both.

A moon landing by two astronauts — Artemis IV — could follow as soon as 2028 using either Starship or Blue Moon, whichever lander is safer and ready first. It will be NASA's first lunar landing with a crew since 1972's Apollo 17. The goal this time is a moon base near the lunar south pole, staffed by astronauts as well as robots.

SpaceX is already taking reservations for private flights to the moon and Mars on Starship.

The world's first space tourist, California businessman Dennis Tito, and his wife signed up 3 1/2 years ago for a flight around the moon. The timing is uncertain.

This week, another wealthy space tourist — Chinese-born bitcoin investor Chun Wang — announced he will fly to Mars on Starship's first interplanetary mission. Wang previously chartered a SpaceX polar flight in a Dragon capsule last year and, along with his hand-picked crew, became the first to orbit above the north and south poles.

No price tag or date was revealed for his Mars cruise.

Copyright 2026 NPR

Stay up to date with our Up First newsletter sent every weekday morning.

MOUNTAIN VIEW, Calif. – Google is changing what it means to Google.

The company this week announced significant changes to its search box — that austere, single-line input field on its homepage that has been the world's most popular entry point into the web for around two-and-a-half decades.

The new version looks similar to the old one-line text box, but it's dynamic, expanding with longer queries. Users can also drop videos, pictures and files into it for what Google calls "multimodal" search.

Behind the scenes, a bigger shift is under way. Google is merging artificial intelligence and traditional web search in a move that Liz Reid, who oversees search at Google, said brings "the best of web and the best of AI together."

Critics say folding AI deeper into search risks further muddying the waters around the provenance of information gleaned from the web, and could take agency away from users. A chatbot is likely to return a summary with only a few links to further information, unlike a web search that returns many pages of links.

But the shift is, in some ways, not surprising, given Silicon Valley's hard pivot toward AI, with Google and others investing billions in the technology and refocusing corporate strategies around it.

For about a year, Google has put "AI Overviews" — short summaries — at the top of some search results. "What we've seen with AI Overviews is that people don't want either just an AI or the web. They want a mix of both," said Reid.

She said she's noticed that users have started to ask longer questions, with more natural language, rather than fragments or key words. "They're asking the question that they really have," Reid said.

For Google, that potentially unlocks new understandings of user intentions. "If you start using more natural language, if you're having a conversation, when you've shifted from researching into buying, you've sort of indicated that. And so we can put better ads because we understand what that is," Reid said.

Google is also introducing agentic functionality to search, so that users can ask it to do tasks over time — like search for theater tickets at regular intervals, or send shoppers a notification when something goes on sale, or conduct a weekly scan of the internet for local events.

Carolina Milanesi, an independent technology analyst, said Google is trying to make its cash cow business — search — richer and more personalized, and it will make shopping easier. But there is a risk that users may have fewer choices about what to click.

"Right now it's: I ask a question, I get a bunch of answers and I feel that I'm in control as to which answer I take, or if I'm looking for something, which product I'm going to end up buying. That is going to be less so going forward," she said.

Milanesi envisions AI-enabled search and agents proposing products to consumers — perhaps even those they have requested — but with less clarity or choice around where it's coming from.

"If you're going to say: 'I want a pair of Jordans, go find them,' you're not necessarily sure what steps have been taken and whether the AI has used a source or a store that was paid for and therefore came up in the search results," she said, "or if AI actually went and did their due diligence and picked the best for me as a customer."

Sarah T. Roberts, director of the Center for Critical Internet Inquiry at UCLA, said the algorithmic underpinnings of Google's web search results have long been "by design, inscrutable to end users" and there's more to it than simply the best of the web floating to the top of any given search. Adding AI will only make the system more opaque, she said.

"What's happening now with AI is that that complexity that already existed will be further obfuscated and even more difficult to unpack," she said.

She noted episodes where Google's AI has provided bad results, including advising putting glue in pizza and eating rocks. "Those gaffes shouldn't be forgotten as Google makes this transition," she said.

And critics say that driving more Google users from web searches to interacting with AI will exacerbate the risks of the so-called "Google Zero" scenario, where the growth of AI queries kills off web search and suffocates the internet click economy as we know it. That includes online shops, web advertisers and news organizations that all depend on referred traffic from Google.

While the redesigned box will be the same for all Google users, there are various tricks and tips online for people who want to disable or avoid some AI functions when using Google.

Google is a financial supporter of NPR.

Copyright 2026 NPR

NPR is dedicating a week to stories and conversations about how communities are moving forward on climate solutions despite significant political headwinds. As the federal government halts plans to address climate change, states, cities, regions and even neighborhoods are trying to fill the gap by cutting climate pollution and adapting to extreme weather.

TACOMA, Wash. — Wendy Clapp dreamt of turning her backyard in Tacoma into a Pacific Northwest forest. But for 25 years, she was stuck fighting back an aggressive invasive species that filled the yard: Japanese knotweed.

Her search to destroy this noxious bamboo-like weed and restore her yard got her thinking.

" What if Tacoma had never been developed?" Clapp questioned.

She imagined her Victorian home as it might have existed hundreds of years ago, with native plants all around.

Clapp turned to a decades-old planting method from Japan to design her forest. The Miyawaki method involves planting native trees and vegetation close together — so densely that 350 trees can fit in an area as small as six parking spaces. The plants compete for nutrients and sunlight, forcing them to grow quickly. Within 20 to 30 years, a fully mature pocket forest emerges.

Clapp's forest begins through a wooden gate.

Wild strawberries, ferns and Pacific ninebark cover the ground.

A big leaf maple stands not far from Clapp's pride and joy, her paper birch.

Clapp's burgeoning pocket-size forest is one of thousands found throughout the world, including India, Ireland, Brazil and the U.S.

The Yakama Nation in Washington state planted a small forest at the Yakama Nation Corrections & Rehabilitation Facility six years ago.

"We live in such a world right now, where we strive to learn about what [undisturbed land] looked like without knowing it," said Marylee Jones, a gatherer and member of the Yakama Nation. "When you do things like this, you're setting up opportunities."

Those opportunities include shade for visitors during hot summer months in the high desert. It's also a place to heal and to learn about traditional plants.

"When you're out here, you start to understand the value of sunshine, and you start to understand the value of not just the tick-tock of the clock, but of how many heartbeats you have in one day," Jones said.

On the other side of the country in May, 50-some volunteers gathered in Attleboro, Mass., to plant a Miyawaki-style forest. Hundreds of trees and plants will turn an abandoned baseball field into an area that can absorb water during heavy rain.

The forests are helping communities adapt to a warming world. But as to whether they can reduce planet-warming pollution? That's less clear. Researchers tell NPR those benefits might be overstated.

A forest born out of industrialization 

The late Japanese ecologist and botanist Akira Miyawaki developed this planting method in the 1970s. He wanted to re-create lush native forests preserved in sacred areas near temples and shrines in Japan.

He imagined a place before people and what vegetation would have naturally existed. This concept of potential natural vegetation helped him select different native species for projects. He then planted a variety of trees tightly together. He found that these forests competed for sunlight and nutrients, which forced them to grow up to 10 times faster than if they had taken root on their own.

Japanese law in the 1970s required industrial companies to have green areas on their sites to prevent pollution and disasters.

Fumito Koike, a professor emeritus specializing in ecology at Yokohama National University, where Miyawaki taught, said Japanese companies asked Miyawaki to plant small, dense forests on their lands.

Miyawaki worked with companies, including Nippon Steel and Mitsubishi, hosting workshops for employees to teach them how to quickly turn barren land into rich, mature forests.

Soon, Miyawaki's work took off. Japanese multinational companies asked him to work on their sites overseas. Miyawaki, who died in 2021, led plantings in 15 different countries, including Malaysia and China.

He claimed that restoring native forests is one of the best ways to prevent further ecological disasters and improve carbon dioxide absorption.

Loading...

Are mini-forests a climate solution? 

"I believe that creating indigenous and real forests, and covering as much of the land as possible with forests, is the most certain and effective measure to reduce carbon dioxide," Miyawaki wrote in his essay "A Call to Plant Trees."

The method has caught on, but whether these pocket forests are the golden ticket for sucking up and storing carbon dioxide isn't clear.

Narkis Morales, who works at the Bioeconomy Science Institute in New Zealand, sifted through 51 published studies about the Miyawaki method. He found that a lot of the claims about the method haven't been verified. He likens it to a placebo in a medical trial.

" Are you going to give someone a drug that perhaps is not going to have any effect on that person?" Morales asked.

His findings, published in the British Ecological Society's Journal of Applied Ecology, show Miyawaki forests haven't been as rigorously evaluated as other planting methods. The study finds most of the claimed benefits, including rapid growth, higher carbon sequestration and increased tree density, don't have enough research to back them up.

It can also be costly to plant a Miyawaki forest. Morales, based on available data, estimated a project in the United Kingdom could cost $1.3 million per hectare. A different method that supports existing natural vegetation rather than planting new trees costs, on average, $143 per hectare.

So for cities wanting to address climate issues, Morales said: "Perhaps there are better options or ways to do it in a more efficient way."

Morales, however, said planting these dense pocket forests is better than cities covered in concrete.

"Any vegetational cover that you put in the city is going to have a positive effect, especially with heat," Morales said. "People have a place to gather or sit or meditate, or kids can go there and play."

Fazal Rashid, an ecological gardener working in India, is also critical of the Miyawaki method after trying it himself. He said in an email to NPR that ecosystems vary and require different approaches.

Rashid said that to effectively rewild the landscape, people need to try different planting tactics in small areas to see what happens.

"Instead of reaching out for a supposedly universal one-size-fits-all insta-formula like the Miyawaki method," Rashid wrote, "we need to accept that rewilding is more about local people reconnecting with their local ecologies and beginning the slow process of restoring a relationship with each other and the land."

Community buy-in matters

In May, high school and college students, along with other community members in Attleboro, Mass., planted 550 native saplings in the once hard-packed soil of an old baseball field.

Now, students hope their efforts will help the community.

"This will cool the area that it surrounds and sponge the water that comes in from flooding," said Jamie Young, a junior at Clark University who helped.

That's a big deal in a place that has dealt with flooding, most recently in 2023.

John Rogan,  a professor of geography at Clark University, helped plant the forest in Attleboro. He has studied the impact of community engagement on tree survivability.

" The best tree planting you could ever have is by people who are trained, who live in that street, and they have ownership of that tree," Rogan said.

In other words, community stewardship is essential for the survival of urban trees.

And it's the community that showed up two years ago to help Wendy Clapp in Tacoma, Wash., plant her backyard mini-forest. They gathered on her birthday for a planting party, then celebrated around a fire later that night.

Clapp has continued to add more plants in recent years.

She said for her, the Miyawaki method has worked — the invasive Japanese knotweed that's long taken over the backyard is now getting crowded out by native plants.

" This is the first time I've seen real hope, where I see, like, we're actually making a difference out here now," Clapp said.

Lauren Gallup is the South Sound Reporter at Northwest Public Radio. Courtney Flatt is a Senior Correspondent at Northwest Public Radio. And, Bianca Garcia is the Climate and Environment Reporting Fellow at WBUR.

Copyright 2026 NPR

Loading...

This story was originally published by CalMatters. Sign up for their newsletters.

Amid tech layoffs, anxiety around artificial intelligence and a forthcoming run for president, Gov. Gavin Newsom today signed an executive order that calls for state agencies to explore ways to mitigate job losses stemming from AI.

The order, among other things, requires state agencies to find ways to help displaced workers, including through severance policies and subsidized employment, and to assess the impact of AI on the California labor market.

It also calls for increased job training , stock compensation, cooperative business ownership for workers and a review of how unions are negotiating over AI.

Newsom signed executive orders last month and in 2023 simultaneously putting in place AI protections and encouraging state agencies to use the technology.

The latest order comes a day after Facebook owner Meta laid off 8,000 workers, with CEO Mark Zuckerberg citing AI in a memo to staff after the cuts. It also comes two days after the California Senate passed the No Robo Bosses Act, which prevents businesses from using decisions made by AI and other automated systems as the sole reason a person gets fired or disciplined. Newsom vetoed a similar bill last fall.

In February, AFL-CIO president Liz Shuler, members of the California Labor Federation and labor leaders in Democratic primary states pledged to pull support for a Newsom 2028 presidential campaign if he didn’t take steps to protect workers from artificial intelligence. Newsom’s veto of the predecessor of the No Robo Bosses Act was named as a reason for that pledge.

This article was originally published on CalMatters and was republished under the Creative Commons Attribution-NonCommercial-NoDerivatives license.

As SpaceX prepares for its public offering, it's undergoing a critical test of the largest rocket ever built.

On Friday, the company is set to conduct a launch of a heavily redesigned version of its Starship rocket after scrubbing a planned launch at the last minute on Thursday. The new rocket will fly using dozens of new Raptor 3 engines powered by a novel fuel system in the booster. It will also carry upgraded avionics, satellites, and test ports for a future refueling system that could allow Starship to one day reach other destinations like the Moon and even Mars.

The launch comes a little under a month before investors expect SpaceX's much-anticipated initial public offering. Many analysts expect SpaceX to raise up to $75 billion, and to be valued at up to $1.5 trillion. The amount raised could make this the largest IPO ever, and make SpaceX one of the most valuable companies in the world.

But a lot of that may rest on how Starship performs.

The launch is "super important for the IPO," said Franco Granda, a senior researcher who covers SpaceX for the data firm, PitchBook. He believes that if Starship's launch goes badly, it could cause investors' excitement for the IPO "to diminish quite dramatically."

"Even though tests are inherently tests, and failure typically doesn't dictate what happens later on, I think SpaceX will want to get this one right," he said.

Financial disclosures released on Wednesday show just how critical the test is. SpaceX is spending billions of dollars per year developing Starship. The development is eating up the profits from its launch business, which operated at a $662 million loss in the first quarter of this year.

Cheap, frequent launches key to the business model

Starship is unlike any rocket ever built. Standing at around 400 feet in height, it's made of durable, but heavy, stainless steel. To overcome its bulk, the spacecraft sits atop an enormous booster called "Super Heavy" that shoots it skyward with 33 Raptor engines. After Starship separates from the booster, it can ignite its six engines to get to space, while the booster is capable of flying back to earth and landing at its launchpad.

SpaceX CEO Elon Musk has said that the ultimate goal is to make Starship return to the pad, allowing the entire system to be reused almost instantaneously. If it works, it would radically lower the cost of launches.

That's key to SpaceX's short-term and long-term vision. The rocket is expected to carry larger, more capable satellites for the company's Starlink internet service. It's also being developed into a lunar lander for NASA and as a possible rocket to Mars. Most recently, Musk proposed that Starship could be used to launch data centers into space, where they could use solar power to run AI chips.

"Starship is a critical piece of the puzzle," Musk said during an event in March to unveil a new chip fabrication facility. To achieve SpaceX's goals, "you need massive payload to space, and Starship will enable that."

The new financial disclosure document reveals just how much of SpaceX's future relies on Starship's development. 

"Any failure or delay in the development of Starship at scale or in achieving the required launch cadence, reusability and capabilities thereafter would delay or limit our ability to execute our growth strategy," the document says. 

It also puts real numbers on how much money SpaceX is spending to develop the giant rocket. In 2025, the company spent a little over $3 billion dollars. In the first quarter of 2026, it spent another $930 million on Starship development.

Success is far from assured. The first version of Starship launched in April of 2023, but failed to separate from its booster and tumbled out of control before exploding. It took three tries before the rocket eventually reached space. The second version of the spacecraft was plagued by multiple failed attempts throughout 2025, though its final two launches went as expected.

This third version is an ambitious redesign, said Scott Manley, an engineer and YouTuber who closely tracks Starship's development. Manley said that the latest version of Starship is more refined. Heat tiles appear more carefully placed, and clunky, temporary fixes, like a "hot staging ring" that was bolted onto the booster after the first accident, have been integrated into the design.

"It looks more like a rocket and less like a bunch of grain silos welded together," he said.

Manly said he's particularly interested in how the rocket's new Raptor 3 engines perform. The engines have been heavily redesigned to increase thrust and eliminate the need for bulky shielding on the bottom of the rocket. The engines have been extensively tested, but they haven't been to space before. "We don't know how they're going to perform under flight circumstances."

Even if the newest version of Starship flies as expected, the space launch company faces a long path towards making the behemoth rocket work as a business proposition. Starship's heat shield has not yet proven durable enough to survive multiple trips through the atmosphere and the spacecraft itself has yet to attempt a landing at the pad in Brownsville, Texas.

Getting all the parts of Starship to work as planned is far more complex than what SpaceX has done with its existing rocket, the Falcon 9, said Tim Farrar, the president of TMF associates, which analyzes mobile satellite services. Starship is "a multidimensional problem that they haven't actually solved yet," he said.

"You can't justify a valuation well in excess of a trillion dollars based on what SpaceX is doing today," Farrar said. "You've got to believe that Musk will come up with something much bigger than that."

Copyright 2026 NPR

Standing in his laboratory, Harvard professor Sean Eddy gazes at a row of vacant work stations. More than a year ago, this lab was filled with over a dozen researchers. On a given day they might be working independently on analyzing genomic sequencing or gathered around the group table, drinking coffee and helping each other troubleshoot questions about genomic data from different species.

Now, after his funding was terminated under the Trump administration, the computer screens are gone and the room is silent. He's one of the last people left.

" Seeing these labs empty — this is not the way it's supposed to be," he says. "This was a very vibrant lab."

Eddy is a computational biologist. He has devoted his career to one fundamental question. " I'm really interested in the origin of life," he says. "I want to know where it all came from."

He and his colleagues spent years developing software that could be used to seek out an answer. Scientists around the world now use the tools his team created to compare DNA and protein sequences, identify genes, and predict what they do. Their work underpins countless studies, including research related to cancer and neurodevelopmental disorders.

It's hard to quantify how much modern science relies on what his team built. Eddy describes its use as being as ubiquitous as microscopes or pipettes.

"It's very affirming for me to pick up sort of semi-random papers in the literature in fields that I care about and see them using our software over and over again," he says.

When the lab was designed more than a decade ago, he worked closely with an architect. On the wall are pictures of animals. His daughter, who was 12 at the time, stenciled them for him. Mixed in with pictures of mice and fish are laboratory creatures. "There's a bacterial virus called T4 that I did my thesis on," he says, pointing to the wall.

In 2025, Eddy received a letter from the National Institutes of Health, informing him that his work "had been determined to be of absolutely no value to the US taxpayer, and therefore it was being specifically terminated," he recalls.

Eddy is one of thousands of researchers across the U.S. still grappling with the damage inflicted on science in 2025 under the Trump administration — despite a restoration of funding earlier this year.

Left guessing 

At the time he received the letter, Eddy had more than a dozen people working for him. Over the last year he's had to let almost all of them go. He's worked closely with them to help find jobs elsewhere.

Eddy says he has given up on any dream that his funding would be restored. "I haven't talked to my program officer in years now," he says. "My guess is that he's under instructions not to talk to me. So we're just sort of left guessing what the status of the grant is."

He estimates the funding loss set him and his lab back by a decade. At 60, Eddy had planned to continue working through the next decade with his team. "For someone of my career stage, this is probably not recoverable," he says.

Walking through the empty lab, he looks at the bare desks where his team used to sit. He'd like to see this lab taken over by a younger computational biologist, someone who could pick up where he left off. But with Harvard now on a hiring freeze, he says, he doesn't see that happening anytime soon.

Money on paper — but not in practice

Champions of science celebrated a rare bipartisan victory in the early months of 2026. After the Trump administration tried to cut, freeze or suspend billions of dollars the previous year, a handful of Republicans — at the urging of their constituents — joined Democratic colleagues in an effort to quietly restore significant portions of that funding through the appropriations process.

Now, many of those same advocates are warning that money is not reaching scientists at the rate it should be, and that a lack of transparency at the agency is compromising the integrity and reliability of its research.

"In the past you had a pretty good sense of how NIH was gonna behave," says Jeremy Berg, a former high-ranking official at NIH who has become a kind of watchdog for the organization. When the Trump administration started slashing funding for NIH, Berg took it personally. " Now that level of trust is pretty much gone," he says.

In the past, says Berg, there was an ethos in the agency that dictated clear deadlines, funding forecasts, and expectations from researchers. This reliability fostered good science, Berg says. He credits the institution with funding and fostering much of the progress in biomedical research in the past few decades — such as mapping the human genome, major advances in cancer care, or new therapies for HIV and AIDS.

Berg recalls something a Republican senator once told him about the agency. "He used to refer to it not just as the crown jewel in biomedical research, but as the crown jewel in the federal government," he says. "I think that one can make a pretty strong case for that."

When the cuts hit, he started tracking their progress, charting the changes over the last year. In 2026, Berg says the budget may look intact on paper. However, he says NIH has switched the strategy to making fewer grants with more money over more years, an accounting shift that means fewer scientists are getting funding.

Berg's analysis showed that at one point earlier this year, NIH had issued roughly 2,300 new grants — about half as many as at the same point the previous year.

"There's a lot of pain and a lot of science that isn't gonna get done," he says.

Advocacy groups have also been sounding alarm bells about a lack of transparency at NIH. Money approved by Congress this year has been slow to reach researchers, they say. Analysis from the Association of American Universities showed that NIH issued 66 percent fewer grant awards in the first few months of 2026 than they did the previous year.

" I'm sadly watching the agency where I worked for so many years be dismantled," says Elizabeth Ginexi, who was a program officer at the agency for 22 years, working on substance use prevention. She left when the Trump administration started making cuts, fearing she would be cut anyway.

She's been looking for a job for over a year.

In the meantime, Ginexi's been analyzing something on the NIH website called forecasts, areas of research the agency would like to fund. Typically these forecasts give direction to scientists who are applying for research money.

Ginexi started tracking them when she observed that they were not being filled as quickly as they had in the past. "There are tons and tons of them — starting from last year — that are still sitting as forecasts and were never published," she explains.

Her research shows that of 336 NIH funding forecasts still listed as open, 205 were already past their promised posting date with no full announcement ever published. It's a way, she says, of giving the illusion of funding opportunities, even as they fail to materialize.

Chances of funding? "Basically zero"

Sitting at her lab, cancer researcher Rachael Sirianni scrolls down the website for the NIH, monitoring the grants she's submitted that are waiting for the agency to review. She looks at one application. " The chances of that grant being funded in 2026 are basically zero," she says.

Sirianni had been counting on that grant to continue evaluating a combination of medications for treating children with cancer that had metastasized to the brain. The drugs together offered a "one-two punch," she says, and was showing a lot of promise. She figured with this progress, she'd be able to secure more funding for her work. But she hasn't been able to see it through the normal review process at NIH.

Many of the families grappling with this condition have no other options for dealing with this kind of pediatric cancer, which is basically impossible to remove or mitigate.

"It's thin and it's across the soft tissues of the brain and spinal cord," she explains. "There isn't really a consistent neurosurgical solution to that cancer complication."

Sirianni is a biomedical engineer. Earlier in her career, while working at a research institute, she met a family who lost a child to a type of cancer considered unsurvivable. "Being exposed to that family's pain, especially when I had become a parent myself," she says, "was pretty personally transformative."

In 2022, she moved her young family from Texas to Worcester, Mass., a city of a little over 200,000 an hour outside Boston, to build a lab at the University of Massachusetts Chan Medical School and run pediatric cancer studies.

For this particular grant, Sirianni worked with a colleague for several years before they submitted their proposal for funding, carefully tracking compliance requirements and watching deadlines. In the last year, these deadlines have been repeatedly moved, making it now impossible for the grant to even be reviewed in time for funding.

Sirianni looks at one laboratory bench that is full of equipment — reagent bottles and pipettes. She's had to lay off the researcher who was working here. "I can't bring myself to clear his bench," she says. "It makes me sad."

In response to her concerns and those laid out by other researchers in this story, a spokesperson from Health and Human Services, Andrew Nixon, acknowledged the slowdown in funding and attributed the delays to the government shutdown and congressional Democrats.

"Timelines have returned to typical funding patterns," he wrote in an email to NPR.

Both Sirianni and Eddy say for them, it's too late to restart their research. "That means that the therapeutic development work that taxpayers previously invested in is now hitting a brick wall," says Sirianni.

"Even as just a citizen of the country, this frustrates me," she says. "It's a loss of investment. It's a loss of momentum for the families that have children that are affected by these tumors. Every month, every week — that matters to them."

Copyright 2026 NPR

Global health emergencies are back in the headlines, with recent outbreaks of hantavirus on a cruise ship and Ebola in the Democratic Republic of Congo.

The internet has responded accordingly, with the situation evoking painful reminders of COVID-19 for many people. Questions filled with fear have surfaced on Reddit, comedic videos are all over TikTok and Instagram, and search terms involving the word "pandemic" have increased on Google Trends in recent weeks.

The COVID-19 pandemic introduced people across the U.S. to a global health emergency that they may have never imagined.

That experience is coloring how some people are thinking about Ebola and hantavirus, public health and infectious disease experts say. Fear around exotic-sounding diseases has always existed, but now people know how a pandemic can change their life.

As Chandra Harvey, a content creator on Instagram whose joking video about another possible pandemic received over 100,000 views, told NPR: "We're all dealing with PTSD from COVID."

For Harvey, COVID-19 "heavily impacted" her family, with a few relatives hospitalized. "COVID scarred all of us," she said.

Early in the COVID-19 pandemic, "you were worried about your friends and neighbors and loved ones dying from COVID," said Dr. Ali S. Khan, dean of the College of Public Health at the University of Nebraska Medical Center. More than 1 million Americans died of COVID-19.

Despite Ebola currently spreading in parts of eastern Africa, infectious disease experts told NPR that the average American should not be concerned about Ebola or hantavirus becoming a repeat of COVID-19.

Here's what to know about how Ebola and hantavirus differ from COVID-19, as well as what people should keep in mind when reading alarming headlines or scrolling through social media.

The COVID-19 effect on Americans 

There is the "dread factor" with certain diseases, said Dr. Amesh Adalja, a senior scholar at the Johns Hopkins Center for Health Security and an infectious disease physician. Even though far more people die from influenza on average every year than from Ebola, "certain diseases spark dread in people," he said.

Harvey, the content creator, said hearing about two diseases back-to-back "jumped out" to her family.

"Ebola … with the combination of hantavirus at the same time, it's just like, it's too much," she said. "Any time you hear of anything from a virus perspective, it's just scary."

Adalja said the memory of COVID-19 also causes Americans to lump outbreaks together.

But the diseases spread differently. COVID-19 (like measles) can spread through the air. Ebola is typically spread through bodily fluids, such as vomit or blood. Hantavirus most often spreads to humans through contact with urine, feces or saliva from infected rodents, though one strain has been identified that can spread from person to person.

"The nuances of the biology of different pathogens, the trajectories of different outbreaks, that all gets lost because what [many people are] worried about is having a disruptive event like COVID upend their entire life," Adalja said.

Some may also be alarmed about how deadly these diseases are and the lack of treatment options, as well as the coincidence that they both gained attention in the same month.

Also, the delay in detecting the Ebola outbreak has made it harder to control the situation, infectious disease specialists told NPR. Dr. Craig Spencer, an associate professor of public health at Brown University and an emergency medicine physician with Brown's Pandemic Center, told NPR's A Martínez that the Trump administration's firing of staff at the Centers for Disease Control and Prevention, the cutting of the U.S. Agency for International Development and the U.S. withdrawal from the World Health Organization are having an impact on the current response in the Democratic Republic of Congo.

In a statement to NPR on Monday, the State Department said it was "false to claim that the USAID reform has negatively impacted our ability to respond to Ebola," adding that funding and support to combat Ebola would continue.

Caitlin Rivers, an epidemiologist at Johns Hopkins, said she's "very concerned about the Ebola outbreak as an epidemiologist."

However, "I'm not worried as a mom, meaning I'm not expecting Ebola to influence my community or really the United States," said Rivers, who wrote the book Crisis Averted: The Hidden Science of Fighting Outbreaks. Similarly, public health officials say the risk to the general public from hantavirus is very low.

The larger context of pandemics 

Pandemics and epidemics have been a part of American and global life for centuries, going back to ancient times.

In the 20th century, flu pandemics dominated the years of 1918, 1957 and 1968. In this century, there have been notable outbreaks of SARS (severe acute respiratory syndrome), the H1N1 flu virus (swine flu), the Zika virus, mpox and measles. Ebola has also had multiple outbreaks in the past few decades, with the one in 2014 killing more than 11,000 people.

Humans have been fighting infectious diseases since they first evolved, Adalja said.

Adalja said he puts the Ebola and hantavirus outbreaks into the larger history of pandemics, epidemics and outbreaks: "Not everything has the ability to be this disruptive force the way COVID was."

"[People] have to understand that pandemics have always been something that humans are going to face," Adalja said. "But science and technology and medicine, they offer us the ability to master these issues, to make them less impactful, to be proactive."

Rivers, the epidemiologist, said that in her experience, a major outbreak of international attention happens about every two years.

"They're a lot more frequent than I think many people appreciate."

What to think about when you hear about a new disease

Instead of immediately turning to dread, experts said, Americans can focus on targeted questions and stay informed with information from local, state and national health officials.

Adalja said to ask: "Are people talking about this spreading the same way COVID does?" Similarly, Dr. Abraar Karan, an infectious disease physician and faculty member at Stanford University, thinks about personal risk: "Are [experts] talking about my individual risk, like me leaving my house and going to work and coming home?"

Adalja also said to "be very wary of what you see on social media if it's not coming from an official press source, because there is a lot of disinformation being deliberately spread."

Harvey, the Instagram influencer, said that since COVID-19, she calls herself the "hand-washing police." She carries hand sanitizer in her purse and constantly makes her son wash his hands. "A lot of those things became routines for us."

Rivers said if people are worried, they can wear a mask and avoid crowded indoor spaces or, if available, vaccinate. "Focusing on those controllables can be helpful," she said.

Copyright 2026 NPR

NPR is dedicating a week to stories and conversations about how local communities are moving forward on climate solutions

LONDON — Until two years ago, West London's Greenford Tube station used to flood whenever it rained heavily. The train tracks are aboveground, but the ticket office would often get inundated. Sandbags still line the corridor.

But in October 2023, a new family moved in nearby, determined to halt the water. The family members built their house from scratch with local wood and kept odd hours, sleeping all day and working only at dawn and dusk. They even put their young children to work.

The new neighbors were beavers.

The beavers are part of an unlikely effort to bring back a vanished species and help Britain adapt to a very modern problem: climate change.

Britain is famous for drizzle, but climate change is making rainfall heavier and more erratic. Places that didn't used to flood are now waterlogged. So scientists have enlisted some of the animal kingdom's best flood engineers — beavers — to help.

In West London, conservationists got a government license to resettle a family of five beavers in a 20-acre urban park near the Greenford Tube station. It used to be a golf course, with a creek running through it. Within weeks, the beavers dammed up the creek, creating a pond that holds water and stops it from spilling into the city. They also diverted the creek's flow into smaller tributaries, creating a wetland that better absorbs heavy rainfall — mitigating the risk of flooding downstream.

Loading...

"They effectively turned this site into a giant sponge that can take heavy rainfall and slowly release water back into the landscape, creating a lot more resilience for flooding," explains Sean McCormack, a local veterinarian who started the Ealing Beaver Project, named for the London borough of Ealing, where it's located.

Not only has the local Tube station stopped flooding, but the beavers have also coaxed back other species.

"By felling trees, they've also opened up the canopy, and we've seen an abundance of biodiversity," McCormack says.

Freshwater shrimp have appeared in the creek, he says, plus eight new species of birds, two types of bats and rare brown hairstreak butterflies, which lay their eggs on blackthorn branches nibbled by beavers.

The beavers have also allowed the city to scrap expensive plans to dig a reservoir and levee.

"We said the beavers can do it for a fraction of the cost, certainly more sustainably," McCormack says.

Now, joggers and teenagers stop to gawk at the beavers in action. There are guided walks and beaver safaris.

On a recent spring evening, a reddish-brown adult scampered in and out of the water, chomping on a felled willow tree. Eurasian beavers can weigh up to 65 pounds; this one was the size of a fat golden retriever.

The Ealing Beaver Project is one of dozens of sites across Britain where land managers are using beavers to restore wetlands and tame flooding.

But first, they had to bring them back from extinction.

Loading...

Reintroducing beavers to Britain for the first time in centuries

In Britain, humans hunted beavers to extinction more than 400 years ago. By the early 20th century, only about 1,200 native beavers were left in Europe and northern Asia, surviving in parts of Norway, France, Germany, Belarus, Ukraine, Russia, Mongolia and China. Sweden reintroduced them in the 1920s, and other countries followed — part of a broader effort to restore native species.

Loading...

By studying fossils, scientists determined that today's Norwegian beavers are genetically most similar to the beavers that lived in Britain centuries ago. So in 2009, wildlife officials relocated two Norwegian beavers to Knapdale Forest, a temperate rainforest in western Scotland. That pair, named Millie and Bjornar, became the Adam and Eve of the modern-day British beaver population. The Scottish forestry department calls them the "original beaver power couple."

"We became kind of attached to Millie and Bjornar," says Pete Creech, a forest ranger who remembers when they arrived, scrambling out of crates and splashing into a loch. He recalls their enthusiasm: "Lots of squeaking!"

Creech set up hidden cameras to capture their crepuscular (dawn and dusk) movements. Within weeks, the beavers dammed up a tiny river, creating an enormous lagoon where swans now nest.

While the United Kingdom overall is getting wetter, some areas — including parts of Scotland — are getting drier, even seeing a growing threat of wildfires. Beavers ensure this rainforest stays wet and, thus, abundant. That's especially important at a time when wetlands are disappearing, with many drained for development.

"Wetlands are one of the most biodiverse habitats in the world," Creech notes. "The U.K. has lost over 95% of its wetlands, and now we're frantically trying to put them back."

Not everyone thinks rodents are the best way to do that, though.

Conflict with farmers

Unlike in London, where they're enclosed in urban parks, beavers in Scotland went forth and multiplied, spreading onto private land. Their numbers have been boosted by beaver bombers — renegade wildlife enthusiasts who've released unlicensed beavers into areas where they might not be welcome.

"As the beaver population has expanded, we've seen more [farmers] getting concerned," says Kate Maitland, a regional representative for Scotland's National Farmers Union.

Beavers can dam up irrigation channels, flooding crops.

"It's quite devastating to see acres and acres of your land sitting underwater," Maitland says.

They can also fell centuries-old trees and collapse riverbanks, exacerbating erosion. Maitland, a farmer's daughter, says she once got the full length of her leg stuck in a beaver burrow while walking along the banks of a stream on her family's land.

The Scottish government has set up a fund to rebuild riverbanks and other beaver damage, if repairs are in the public interest. That doesn't typically cover damage to private land.

Some farmers shoot beavers, though they need a license to do so, since beavers are a protected species. It's also illegal to disassemble beaver dams or lodges that are more than two weeks old. Instead, farmers are encouraged to call wildlife officials, who trap and relocate beavers. That's where London's beavers came from.

Other farmers have learned to like the new neighbors — and even celebrate them.

Learning to live with beavers

Tom Bowser is a fifth-generation farmer in central Scotland. He has empathy for his fellow farmers: "When you're trying to grow food, the presence of a fat semiaquatic rodent who wants to raise water levels is understandably going to be unpopular!"

Bowser's farm is strewn with trees felled by beavers. Many of the tree trunks have been whittled into hourglass shapes by beavers' sharp teeth. Some of them teeter, about to fall.

He finds it fascinating.

Bowser wraps young trees in chicken wire if he wants to protect them. (In London, officials painted trunks with sand, which gets stuck in beavers' teeth.) But he has found that the benefits outweigh the costs.

A beaver dam has diverted floodwaters from his driveway, creating a pond lined with benches that's frequented by tourists. He runs spring and summer beaver-watching tours that are especially popular with children, who previously knew beavers only from fairy tales.

"We get people from all around the world coming here now!" Bowser says. "Growing up here, you didn't see any car you didn't recognize."

Beaver fever is spreading

The beaver buzz is catching. The animals have made comebacks in Italy, Portugal and the Ukrainian part of the Danube River delta. In the United States, the Methow Beaver Project releases them into fire-damaged areas of Washington state. In Idaho, NASA is helping track beavers' work.

In Britain, beavers are especially popular with land managers who are short-staffed.

Want more stories like this? Sign up for NPR's Up First Newsletter

South Norwood Country Park is a 125-acre nature reserve with only one employee. Volunteers do some of the ground maintenance. They even don waders to dredge streams once a year.

"That's exactly the sort of work the beavers would do naturally," says countryside warden Ian Glover. He has applied for a license and hopes to welcome beavers in 2028 or 2029.

Like Ealing, South Norwood is on London's urban periphery. It's famous for birds, with boxes for kestrels to nest in perched atop poplars. The park's peak bird count — 177 species — goes back to 1935. But birds have been in decline across Europe.

Glover hopes beavers might help reverse that locally, by damming up streams and creating wetlands that attract more birds.

Beavers build dams and raise water levels in part to hide from predators, Glover notes. But most of their predators — including wolves and bears — have been extinct in Britain for centuries.

"Obviously they haven't gotten the memo," Glover laughs.

And these beavers have been so useful, nobody's telling them.

Loading...

Edited by Rachel Waldholz

Copyright 2026 NPR

NPR is dedicating a week to stories and conversations about how communities are moving forward on climate solutions despite significant political headwinds. As the federal government halts plans to address climate change, states, cities, regions, and even neighborhoods are trying to fill the gap by cutting climate pollution and adapting to extreme weather. 

Lawns are great surfaces for throwing a football or laying out a picnic.

But turfgrass in the United States now covers an estimated 40 million acres — an area about the size of the state of Georgia — and these manicured lawns take an environmental toll.

Gas-powered yard equipment, like lawn mowers and trimmers, put out 30 million tons of air pollutants a year, the Environmental Protection Agency says.

To keep our lawns tidy and green, we also use weedkillers and fertilizers. When it rains, those chemicals escape our yards and make their way through storm drains to our lakes and rivers.

Shrinking lawns can reduce all these impacts. It can also create space for gardens that feed wildlife — great news at a time when North America has lost one-quarter of its birds and the U.S. has lost one-fifth of its butterflies.

University of Delaware entomologist Doug Tallamy has calculated that if Americans reduced their lawns by half and added native plant gardens to feed birds, butterflies and other wildlife, this would create more habitat than Yellowstone and a dozen other major national parks combined.

Interested in ditching your lawn? These tips can get you started.

1. Choose a spot where you'll kill some grass

Consider starting small, especially if you're new to gardening.

" It can be overwhelming to take on an expansive garden," says Stacia Stelk, executive director of Deep Roots KC, a group that teaches the public how and why to plant habitat gardens in the Kansas City region. " As you get more comfortable, there's always room to expand."

Replacing the whole lawn at once can lead to weed, mud and erosion problems, so it requires careful planning and more work.

A good place to remove some grass could be along a fence or sidewalk. If you have a tree in your yard, consider putting a flower bed around it. This is called a soft landing because it gives caterpillars a safe place to go after they finish feeding on your tree's leaves. There, they can make their chrysalises and cocoons among flowers and leaf litter, safe from lawnmowers, and emerge as adult butterflies and moths. Bees, fireflies and other insects will find homes there too.

2. Plan what you are going to plant

Using flowers, trees and shrubs that have existed in North America for millennia is a surefire way to turn your space into a wildlife magnet.

That's because many insects feed on specific native plants. A famous example is the monarch caterpillar's need for milkweed — it can't eat anything else.Native plants also boost insect populations, which feed birds, frogs, lizards and other animals.

To pick your specific plants, first note whether your new flower bed will be in shade or sun and what the ground is like. For example, is the soil very sandy? Is the area usually wet?

Next, find a native plant group specific to your region, because native plants for Arizona and New York are completely different. Many of these groups have online cheat sheets for picking plants. Midwesterners, for example, can check out GrowNative.org, which has sample flower bed plans and Top 10 lists of plants for different light and soil conditions.

Also know what style you're looking for. Are you looking for short groundcovers? Do you dream of a cottagecore garden with big drifts of pastel flowers? Do you want a hedge between you and your neighbors?

Finally, know your local rules: Some homeowners associations or city codes restrict tall flowers and grasses in the front yard, for example.

3. Kill your grass

There are plenty of ways to kill your grass. You can kill it by covering it with cardboard to deprive it of sunlight. You can lay down sheets of clear plastic to block light and rain while baking the grass in the sun's heat. You can rent a sod cutter or use a shovel to dig the turf out by hand. You can also use herbicide. Iowa State University has a nice guide to carrying out these options.

There's no right way to kill your grass. The best option for you will likely depend on your particular outdoor space, how much hands-on time you want to spend removing the grass and how long you're willing to wait for it to die. Smothering the grass with cardboard can take weeks or months, for example.

If you're using plastic or cardboard, there's no need to remove the grass once it's completely dead, unless you see seedheads or grass pieces that could resprout. The dead turf will decompose on its own and add organic matter to your garden.

4. Buy plants for your new garden (or look for free options)

The tricky thing about native plants is that typical garden centers often don't stock many of them. Many opt to focus on popular ornamentals such as peonies and boxwoods that originally came from other continents.

Native plant or wildlife advocacy groups often post indexes online of native plant nurseries or pop-up events where native plants will be sold.

But buying lots of plants can be expensive.

Jeffrey Popp, director of restoration at Anne Arundel Watershed Stewards Academy near Annapolis, Md., has a few tips to save money when buying plants.

First, look for places that sell plugs — small, young plants in six-packs or flats.

"You can buy native plant plugs relatively inexpensive," Popp says.

Despite being small, those plugs can grow fast. Also, some homeowners with native plant gardens will give away seedlings for free or swap plants, he says.

Try searching Facebook for local groups dedicated to native plants. This might mean a native plant society, for example, or a local chapter of Wild Ones, a group that encourages wildlife friendly gardening.

You can also start plants from seed, but this takes patience. Nurseries like Prairie Moon, a major native plant seller based in Minnesota, have detailed instructions for how to get each kind of seed to germinate.

5. Plant the plants and wait

Get them in the ground and put mulch around them. There's no need to add extra garden soil during planting if you've picked native plants suited to your soil conditions. Advice on how far to space the plants out varies depending on the plant. But basically, you'll want to space them so that once the plants mature, your flower bed is full.

" Plants want to touch each other," says Paula Diaz, a master gardener in Kansas City. "They don't want acres of mulch in between them."

This is better for you, too, because eventually you won't have to replenish mulch or pull weeds as often.

It'll take a few years to get to that point, though. Gardeners like to say that many perennials "sleep, creep, leap," meaning they won't seem to do much in their first year, will only grow a little in their second, and then will thrive in their third. (Although it's true that some species can grow faster.)

6. Watch out for weeds and dry conditions. And be sure to enjoy your new garden

As young plants grow, you'll especially have to keep an eye out for weeds and to make sure that empty spaces between plants are mulched.

You may also need to water them while they develop strong root systems. If you've chosen plants appropriate to your area, you shouldn't have to water them once they mature, except perhaps during serious dry spells.

In the winter, don't cut the plants all the way to the ground. Leave at least part of the stems because native bees nest in them. Also leave fallen leaves, because butterflies and other critters are overwintering in there.

Enjoy! Diaz has gardened with native plants for more than a decade and five of her neighbors have followed suit.

"There's always a bird that's singing or frogs that are croaking," she says. "Being able to go outside and just walk around and see life that happened because you planted what you planted — it helps your heart."

Celia Llopis-Jepsen is host of the environmental podcast Up From Dust and a reporter for Harvest Public Media.

The podcast episode was produced by Sylvie Douglis. This story was edited by Shahla Farzan, Malaka Gharib and Neela Banerjee. The visual editor is CJ Riculan.

We'd love to hear from you. Email us at LifeKit@npr.org. Listen to Life Kit on Apple Podcasts and Spotify, or sign up for our newsletter.

Copyright 2026 NPR

Glendale Community College's commencement ceremonies hit a snag just as students were walking across the stage to get their diplomas last week. The wrong names were being read aloud at the ceremony, just outside Phoenix. Some of the graduates' names didn't even get read.

The college's president, Tiffany Hernandez, tried to explain the problem. "We're using a new AI system as our reader," she said, leading to loud boos from the audience. (In a statement, the college blamed technical issues and said it had apologized to students for the experience.)

Other commencement speakers who have brought up the sweeping changes that artificial intelligence is driving are also facing boos from the Class of 2026.

Real estate executive Gloria Caulfield described AI to the graduating class of the University of Central Florida on May 8 as "the next industrial revolution."

The boos started almost immediately.

"OK, I struck a chord," said Caulfield.

Graduating students at Middle Tennessee State University booed when record executive Scott Borchetta told them at their May 9 commencement ceremony, "AI is rewriting production as we sit here." Borschetta responded to the boos with: "Deal with it. Like I said, it's a tool." As the booing continued, he added, "Then do something about it. It's a tool. Make it work for you."

Former Google CEO Eric Schmidt was repeatedly booed by University of Arizona graduates at their commencement on May 15, including when he said, "The question is not whether AI will shape the world. It will. The question is whether you will help shape artificial intelligence."

ChatGPT was released in 2022, when many of this year's undergraduates were just starting college. Many have embraced AI for good and for ill, whether to build businesses or use it to cheat.

But despite – or perhaps because of - those experiences, many graduates feel those boos are justified.

"I think my gut reaction was I would be one of those people in the crowd booing," said Maggie Simmons, who will attend her own graduation ceremony at the University of Denver next month.

She told NPR she is concerned AI is hurting the planet and harming Black and minority communities. AI language models have been found to reinforce systemic racism and data centers needed to power AI systems have had a disproportionate impact on minority neighborhoods.

"The future should be these people in this room that are earning their degree and now going out into the workforce," said Simmons, who studied molecular biology and Spanish to prepare to become a pediatrician one day. "We should be celebrating them and their brains, not some artificial intelligence that in the future is going to take their jobs and especially without regulation."

Kareen Gill, a recent graduate of American University with a political science degree thinks a lot of her generation is feeling pessimistic about AI.

"I think at the beginning we were excited about it and it was this cool thing, 'Oh, I can write an essay for you,' but now like, we don't want that anymore and we don't want it messing with our job prospects and messing with the jobs that we've worked for years — so hard for four years — to kind of be eligible for," Gill said.

One immediate impact Gill said she has noticed is fewer internships and entry level positions doing things like answering phones because AI is replacing some of those jobs.

"So we're seeing that firsthand and we're seeing how much it's disadvantaging us," said Gill. "But I don't think that older generations are necessarily in our shoes in that way. It's not really going to impact their future on the rest of their adulthood in the same way."

Indeed, a March poll from Quinnipiac University showed that there are generational differences in how concerned Americans are about AI taking jobs

"Gen Z, despite being more familiar with AI, is the most pessimistic about jobs, with 81% saying that AI will decrease job opportunities," said Chetan Jaiswal, an associate professor of computer science and associate chair of the Department of Computing at Quinnipiac who also worked on the poll.

Jaiswal said that the poll showed that Americans overall are more concerned and less excited about AI as the technology's impacts are becoming more evident.

"People are not rejecting AI, but people are asking questions now since the initial AI fever is gone," Jaiswal said.

That point was echoed by Gill, the recent AU graduate, who said her generation's concerns about AI go far beyond getting their first jobs.

"How they're making billionaires richer and depleting our environment has really opened our eyes to the ripple effects of AI," she said.

Indeed, Quinnipiac's poll found only 5% of Americans feel AI development is being led by people or organizations that represent their interests.

Copyright 2026 NPR

Updated May 20, 2026 at 3:42 PM PDT

Elon Musk's SpaceX is getting ready to break new records on Wall Street — and make one of the world's wealthiest men even wealthier.

The space-launch company filed financial information with regulators to pull off what could be the biggest initial public offering ever in mid-June. Though it did not disclose how much it intends to raise, previous media reports have said the company is seeking about $80 billion.

That would make it by far the largest IPO in history, much higher than the $29 billion that Saudi Aramco eventually raised after its 2019 offering.

The deal could also value the company at over $1 trillion, making SpaceX one of the most valuable companies in the world, and potentially even worth more than Musk's electric-vehicle company, Tesla.

The IPO would also massively enrich Musk, who's already one of the world's wealthiest people and who could become its first trillionaire. The former adviser to President Trump controls 85% of voting power in the company, according to the documents filed on Wednesday.

Loading...

SpaceX's IPO is a big AI bellwether

SpaceX's initial public offering has been expected for months. Earlier this year, the company filed confidential paperwork with the U.S. Securities and Exchange Commission to start the process of selling its shares to the public, the Associated Press and others reported in April.

The IPO could set off a blockbuster year for Wall Street, which is also anticipating potential listings from OpenAI and Anthropic, the makers of ChatGPT and Claude, respectively. Such IPOs will put a public value on — and eventually allow ordinary investors to buy into — some of the most powerful private tech companies driving the AI boom.

SpaceX is among them. The company has already cornered the market on space launches, makes reusable rockets, and has built the world's only satellite internet service, Starlink. Last year, it merged with Musk's artificial intelligence company, xAI. Going forward, Musk wants to build orbital data centers, moon bases, and eventually put humans on the surface of Mars.

The filing show SpaceX is spending eye-watering amounts of money to fulfill those ambitions. Last year, xAI more than doubled capital spending to $12.7 billion dollars, driven by data center development and construction. The company spent another $3 billion last year developing its massive new rocket, Starship.

In the first quarter of this year, two of SpaceX's three main businesses lost money. Its rocket-launching business reported an operating loss of $662 million, while its AI business lost nearly $2.5 billion. Only its satellite communications business, which runs Starlink, turned a operating profit of $1.2 billion. Overall, the company reported a net loss of nearly $4.3 billion between January and March on total revenue of $4.7 billion.

The pros and cons of IPOs for companies — and ordinary investors

IPOs allow public-markets investors, including individual investors, to buy shares in some of the buzziest and high-profile companies. The deals also raise tremendous amounts of money for the companies that are selling these shares, and the private investors who have funded their pre-IPO growth.

Analysts are also hoping that the SpaceX IPO will kickstart what has been a relatively slow market for initial public offerings in the past several years.

But there are some risks for individual investors looking to buy into SpaceX as it goes public. Post-IPO companies tend to see their share prices underperform the broader market.

"Historically speaking … it's pretty jarring how bad it is," says Franco Granda, a research analyst who covers SpaceX for the data firm PitchBook.

He adds that for closely watched companies like SpaceX, which saw its valuation soar under private ownership, going public also means being much more open to scrutiny from financial regulators and the general public.

"Now they're under much heavier scrutiny, and the [big] valuation sometimes is hard to justify," he says.

SpaceX's public value is closely tied to Elon Musk

Perhaps no company is more closely tied to Elon Musk than SpaceX. Musk founded the company in 2002, with the goal of one day creating a self-sustaining colony on Mars. He single-handedly revolutionized the rocket industry by creating a reusable first-stage that saved huge amounts of money.

SpaceX has gone on to win numerous contracts launching defense satellites, commercial payloads and even astronauts on board its Dragon capsule. It's secured billions in contracts from the Department of Defense and NASA. According to an analysis by the independent firm Brycetech, 85% of all launches last year were aboard SpaceX rocket, putting the private company ahead of even nation states like China and Russia.

Despite this dominance, launching things into space actually accounts for only a small part of SpaceX's revenue, said Tim Farrar, president of TMF associates. By far, the larger part of the company's revenue comes from Starlink; the company generated $11.4bn in revenue from its connectivity division in 2025, compared with $4.1bn from its space launch division and $3.2bn from its AI business, the filings show.

But Farrar thinks even Starlink can't entirely justify a trillion-dollar valuation.

"The valuation is completely dependent on the degree to which people believe in Elon Musk," he said. "It's not dependent on the current business."

Copyright 2026 NPR

Stay up to date with our Up First newsletter sent every weekday morning.

The tech company Meta kicked off a sweeping reorganization on Wednesday that will shrink its workforce and accelerate a pivot toward artificial intelligence.

In an internal memo last month, the parent company of Facebook, Instagram and WhatsApp said it would lay off around 10% of its workforce — or about 8,000 people — in May. Company spokesperson Erica Sackin on Wednesday confirmed to NPR that those affected employees have been notified.

Another 7,000 Meta employees will see their roles change as part of the AI pivot, according to a source familiar with the situation but not authorized to speak publicly about the topic.

"To focus more on AI they are moving 7,000 people to teams that focus on AI projects," the person said.

Sackin declined to comment on what teams were growing or shrinking as a result of the shakeup.

The re-assignments were first reported by Reuters, which cited an internal memo saying the employees would be shifted to four new teams building AI tools and apps. NPR has not independently verified the contents of that report.

Meta and other big tech companies have been placing huge bets on artificial intelligence, luring talent with giant pay packages and building multibillion dollar data centers to try to win the AI race — one in which Meta lags behind competitors such as OpenAI, Anthropic and Google.

In January, Meta forecast capital expenditures this year that are almost double what it spent last year.

Meta's reorganization is part of a string of announcements across the tech sector in recent months of job cuts amid intensified focus and spending on AI.

It also comes in the middle of a rough stretch for the company. It has been pulling back from the virtual reality "Metaverse" that CEO Mark Zuckerberg touted as the company's future when he launched it in 2021.

Earlier this year, Meta also lost pivotal court cases in New Mexico and California claiming that its platforms have been harmful to children and young people's mental health. In June, Meta — among other tech companies — is scheduled to head back to court to face school districts that sued social media companies over claims they caused a costly mental health and social media addiction crisis among students.

Copyright 2026 NPR

When Kendal Sager lifts the top of her beehive, tens of thousands of bees waggle across the honeycomb — their cells filled with bright yellow, orange and pink pollen collected from flowers in Sager's neighborhood.

Sager, a California master beekeeper, said bees do far more than produce that sweet substance: they pollinate crops for hundreds of types of nuts, fruits and vegetables.

"Even if you don't like bees themselves, you have bees to thank for the food on your table," she said.

This Wednesday is World Bee Day, which was established by the United Nations in 2018 to raise awareness about the need to protect the insects. So this installment of NPR's Word of the Week looks at the mystery behind bees' buzzy name — and why they need more of our attention than ever.

Why the origin of 'bee' is difficult to trace

Doug Harper, founder and editor of Etymonline.com, an etymology dictionary, said that unlike many words, the meaning of "bee" has pretty much remained the same over time.

"A word like 'bee' has always been 'bee,' as far back as you can trace it," he said.

According to Harper, "bee" came from the Old English "beo." Some poets in the past would use the term "beowulf", meaning "a wolf to bees," to describe bears.

But the exact origin of "bee" itself remains a linguistic mystery largely because of the insect's long history on the planet, Harper said. Bees have existed for millions of years, and beekeeping has been a practice for over a millennia.

"The words that have been here forever, we'll never know probably," Harper said.

One theory is that it relates to the buzzing sound that the insect makes, he added. Throughout history, other words have been used to refer to the pollinator, like the Latin word "apis" which led to "apiary," or the places where beehives are kept.

But "bee" endured like none other. It now belongs to an array of phrases and idioms: making a "beeline," being a "busy bee," minding your own beeswax. That's a testament to how humans have closely watched and appreciated bees for centuries, according to Harper.

" When I look at words like that, they don't even look like words. They look like landscape. They're so ancient," he said.  "Those things are sort of rare and valuable just for that."

The bees are not all right 

Over the years, it's become harder to keep bee colonies alive. That was most apparent last year, when commercial beekeepers in the U.S. reported losing nearly 56% of their honeybee colonies — the largest decline since the annual U.S. beekeeping survey began in 2010.

"The worry is that at some point they won't be able to keep up anymore," said Mateo Kaiser, a fifth-generation beekeeper in California and the managing director of Swarmed, a resource for beekeepers.

There's a number of factors behind last year's huge collapse, including habitat loss, exposure to harsh pesticides and a tiny, but destructive, parasite known as the Varroa mite, which arrived in the U.S. from Asia in the 1980s.

More broadly, bees are extremely sensitive to changes in the climate, according to Kaiser. If the spring season is too dry, flowers and trees cut back on nectar production, causing bees to go hungry. If there's a sudden cold snap in the spring, it could set bees back in their progress for the whole season, Kaiser added.

"As the climate changes and becomes less predictable, it becomes harder for beekeepers to know what to expect and to make sure that they're taking the best possible care of their bees," he said.

Each year, bees make millions of trips across the country. Their travels not only pollinate crops and produce honey, but also provide insight into how the environment is faring.

"Anything the bees are coming into contact with, it's affecting them and it's telling us something about the health of our communities, and how local biodiversity and ecosystems are being impacted," he said.

As a result, Sager said bees have made her more attuned to the weather and when flowers are blooming. That's one more way that bees are making an impact on humans.

" So even if you don't care about the bees," she said, "it's pointing at a lot of other issues that may cause problems for everyone and other species."

Copyright 2026 NPR

Minnesota Gov. Tim Walz has signed the nation's first law banning prediction market sites from operating in the state, and in response, the Trump administration has sued, teeing up a legal battle over the most far-reaching crackdown on popular services like Kalshi and Polymarket.

It comes as states confront a growing standoff with the Trump administration over how to regulate the industry, which allows people to bet on virtually anything.

The new state law makes it a crime to host or advertise a prediction market, which it defines as a system that lets consumers place a wager on a future outcome, like sports, elections, live entertainment, someone's word choice and world affairs.

The prohibition extends to services supporting prediction markets, like virtual private networks, that could allow consumers to disguise their location and get around the ban.

It would force prediction market sites like Kalshi and Polymarket to leave the state, or face possible felony charges. The law takes effect in August.

"We as a state should decide how best and what regulations we think should attach to gambling, to protect public safety, to protect our kids," said Minnesota Rep. Emma Greenman, the Democrat who introduced the measure.

The law has a carve-out for event contracts that serve as an insurance policy in the event of "harm, or loss sustained" and for the purchase of securities and other commodities.

The Commodity Futures Trading Commission's lawsuit seeks to block the law before it starts, arguing the prediction market industry should be exclusively regulated by federal officials.

"This Minnesota law turns lawful operators and participants in prediction markets into felons overnight," said CFTC Chairman Michael Selig. "Minnesota farmers have relied on critical hedging products on weather and crop-related events for decades to mitigate their risks. Governor Walz chose to put special interests first and American farmers and innovators last."

An updated version of the prediction market bill allows trading on weather, an exception that followed pushback from the agricultural industry, which has historically used futures trading on weather as a hedge against storms and other inclement weather that can affect a harvest. Walz is expected to sign it soon.

Besides Minnesota, bills cracking down on the prediction market industry have been introduced in 14 other states, according to the National Conference of State Legislators. Two of those states, Hawaii and North Carolina, have pending bills seeking to ban the industry statewide.

Experts say the cloud of legal uncertainty hanging over prediction markets apps have not slowed their rapid growth.

"The states are using any tactic they can to go after the prediction market companies," said Melinda Roth, a professor at Washington and Lee University's School of Law, who studies the industry. "But they've embarked on a too big to fail strategy and have become quite mainstream," she said. "It will be hard to put that genie back in the bottle."

A legal fight over the Minnesota ban is expected. Questions over whether states or the federal government should oversee the prediction market industry have already triggered more than 20 lawsuits. One of those cases, in Nevada, led to Kalshi pausing its sports betting in the state after a judge found it "indistinguishable" from state-regulated sports gambling.

The Commodity Futures Trading Commission has filed federal lawsuits against five states, including Arizona, Wisconsin and New York, attempting to override state regulators' attempts to rein in the betting sites.

The CFTC has argued it has exclusive jurisdiction over prediction markets, even though former CFTC members and legal experts say bets on football games, words President Trump might say during a press conference and whether Ricky Martin will make an appearance at the Super Bowl are matters far outside its traditional scope.

In a statement to NPR, Kalshi spokeswoman Elisabeth Diana said banning prediction markets is a "blatant violation" of the law.

"Minnesota banning prediction markets is like trying to ban the New York Stock Exchange," said Diana, adding that "this actively harms users because it reduces competition and drives activity offshore."

A Polymarket spokesman told NPR that Minnesota's ban runs counter to the federal government's "established framework" for regulating prediction markets.

Tribal-owned casinos operate in Minnesota, but online gambling and sports betting are not legal in the state.

Prediction markets like Kalshi and Polymarket have given access to sports betting to people in states where the activity is prohibited, since the Trump administration regulates the sites as a type of "event contract," rather than gambling, which typically is overseen by state gaming authorities.

Nonetheless, sports gambling powers the sites. On Kalshi, for instance, more than 85% of trading activity is related to a sporting event, some of those trades being "parlays," high-risk wagers that multiple things, points scored, fouls, passes, will all happen.

Bettors on the sites are making billions of dollars in trades every week, even as questions around insider trading and how the markets can create perverse incentives for people to manipulate real world outcomes continue to vex the companies.

Minnesota Public Radio News reporters Dana Ferguson and Peter Cox contributed reporting to this story.

Copyright 2026 NPR

Suddenly, a behavior you spent weeks working on is just wiped out. You did everything right — so what went wrong?

For a habit to stick, you need to do more than just repeat it, says behavioral scientist BJ Fogg, founder of Stanford's Behavior Design Lab. You need motivation, the ability to do the action and a prompt to do it. "When those three things come together, the behavior happens."

Fogg shares insights based on the Tiny Habits method, a research-backed approach he developed in 2007 on how to create lasting change.

This comic was drawn by the cartoonist Vreni Stollberger, based on writing by Mika Ellison and reporting by Marielle Segarra.

Vreni Stollberger is a cartoonist based in Queens, N.Y., who loves using comics to break down complex topics, from policy and science to history and culture.

The podcast episode was produced by Sylvie Douglis. The comic was edited by Malaka Gharib. The visual editor is CJ Riculan. We'd love to hear from you. Email us at LifeKit@npr.org. Listen to Life Kit on Apple Podcasts and Spotify, or sign up for our newsletter.

Copyright 2026 NPR