Ross Andersen | The Atlantic

The Most Beautiful Moment of the Artemis II Mission

2026-04-07T16:54:01-04:00

The most moving image to emerge from the Artemis II mission has not been a snapshot of the moon or the Earth. The camera was instead pointed at the astronauts themselves, squeezed inside their tiny capsule. Christina Koch sat in the foreground, strapped into her chair. Only parts of the other three were visible. Jeremy Hansen, the Canadian, was talking to ground control but also to an international livestream audience.

Hansen said that the crew had spent part of yesterday morning looking out the window at the moon. The astronauts had seen an abundance of craters, including a few scars likely incurred about 4 billion years ago, when, during their shared childhood, the Earth and its satellite were both bombarded by asteroids. Many of the lunar dimples and round basins already have official names, but not all of them. Hansen said that the crew would like to propose a couple of new ones.

Naming is a poetic act, and it can go wrong. Before Richard Nixon’s 1972 announcement of a new spacecraft that would carry Americans to orbit more regularly, Peter Flanigan, his assistant, made the case that it ought to have an exciting name. Someone had suggested Pegasus. Naming the program for a winged horse—a working animal that could fly to and fro—made sense, and it was a callback to the classical Greek grandeur of Gemini and Apollo. Flanigan liked “Space Clipper” and “Starlighter,” but he warned against “the Space Shuttle,” Nixon’s eventual choice, because to him, that name connoted “second-class travel.” By emphasizing the routine nature of the cosmic jaunts that the new spacecraft would enable, it risked reminding people of their dreary commute. It robbed the shuttle’s destination—the celestial realm!—of mystique.

For this mission that has just flown around the moon, and those that will succeed it, NASA picked a much more inspired name, better even than the one given to the agency’s previous moon program, more than half a century ago. “Apollo” was never quite right. It is the name of a sun god, an avatar of reason, order, and harmony. Artemis is a proper moon deity. As a wild forest huntress, she embodies the dreamier lunar qualities, the nighttime longing and magic.

On Monday, while flying around the moon, the crew tried to live up to this elevated standard of naming. During the livestream, Hansen said that the crew hoped that a crater on the moon’s far side might share the name of their spacecraft, Integrity. You can understand why they might have been feeling gratitude for the little vessel at that moment. In carrying them farther from Earth than any humans had ever traveled, it had bested the Santa María, the H.M.S. Endeavour, and every single one of the Apollo crew modules. For days, its thin walls had been the only thing separating their soft animal bodies from the lethal vacuum of space.

[Read: Why doesn’t anybody realize we’re going back to the moon?]

Hansen said that the second crater was especially meaningful to the crew. It was located close to the boundary line between the moon’s near and far sides, and can be seen from Earth for part of the year. Hansen proposed that it be named for a departed loved one from their “astronaut family.” To his right was Reid Wiseman, the mission’s commander, who in 2020 lost his wife, Carroll, to a five-year battle with cancer. The couple’s two daughters were teenagers at the time, and since then, he has raised them on his own. “We would like to call it Carroll,” Hansen said of the crater. His voice cracked as he spelled it out. C-A-R-R-O-L-L. The astronauts wiped away tears, and all four of them floated up to the top of the capsule, in a group hug—an image of human tenderness, beamed down to a planet that badly needed one.

Wiseman is now on his way home to his daughters. The crew blasted off from the Atlantic coast, but on Friday, they will splash down in the Pacific. They’ll don entry suits and point Integrity’s heat shield at Earth’s fast-approaching atmosphere. The friction and burn will surround them in a placenta of superheated plasma. When it nears 3,000 degrees Fahrenheit, the astronauts will lose contact with ground control. Parachutes will shoot out of the spacecraft to slow it down and stabilize it. According to NASA, the fabric will have been packed tight, to the density of oak wood. The capsule will splash down off the coast of San Diego, and orange airbags will inflate to flip it upright. Divers from the U.S. Navy will approach in choppers and quickly set up a platform. Someone will slide open the capsule’s door, and the astronauts will come out and huff down sweet lungfuls of sea air.

Why Doesn’t Anybody Realize We’re Going Back to the Moon?

2026-04-01T19:52:00-04:00

The most momentous launch since the Apollo era was about to begin, and along Florida’s space coast, a secondhand exhilaration was working its way through the assembled crowd, as though all of us, and not just the astronauts, would soon ride out of Earth’s gravity well on a pillar of fire. The space faithful had started arriving at the A. Max Brewer Bridge in Titusville before dawn, under the light of a full, yellow moon. They had set up their folding chairs and tripods at the high point of the bridge, to get the best line of sight, and stayed fixed in place during a brief rain, and again later, when a concerning wind blew across the lagoon.

On the launchpad, the rocket stood more than 30 stories high, illuminated by banks of stadium lights. As the day wore on, a tailgate atmosphere took hold in the park at the foot of the bridge. For miles, cars had pulled onto nearly every available stretch of grass. Families spread out on picnic blankets, and college-age kids wheeled out coolers. They threw footballs and baseballs back and forth and drank beer.

Two hours before sunset, thousands of us massed on the bridge. Again and again, people told me that they’d come to see history. If Artemis II is successful, the astronauts will be the first humans to reach the moon’s orbit in more than 50 years, and their path around its far side will take them farther into the universe than any human being has previously traveled. The man next to me was streaming the NASA telecast on his phone. He told me that we were one minute from liftoff.

The Artemis II mission has seemingly come out of nowhere. In January, a few weeks before an earlier launch attempt, I’d started asking friends if they were looking forward to it. Few of them had any idea it was happening. “We’re going back to the moon?” they would ask, with the sort of mild surprise that one might experience upon being told that the Super Bowl is only a week away. They didn’t linger on the subject. Anyway, how are the kids?

The original moon missions of the 1960s could be sold as history’s grandest adventure. The Apollo program was the triple-back-handspring exclamation mark on a century of American technological transformations, during which Americans had electrified their cities, filled their streets with cars and their skies with airplanes, split atoms, and invented digital computers. To complete this phase shift into the future, the country banded together to build a spaceship that carried humans to another world, and performed a flag planting for the ages, a peaceful Iwo Jima. Wernher von Braun, the (ex-Nazi and) intellectual architect of Apollo, had compared the moon landing to the epochal moment when aquatic life had first crawled onto land.

In the decades since, the Apollo program has lost some of its aura, in part because it did not lead to a glorious space future in the way that its architects, including von Braun himself, had hoped. Six decades have passed, during which time we’ve had the invention of the internet, the smartphone, and powerful artificial intelligence, and yet this year’s launch is not to Mars or the outer planets or the stars. We’re just returning to the moon. Even though three-fourths of the world’s population is young enough to have never experienced a crewed lunar mission like this one, it has the feel of a rerun. Again we’re rushing to arrive before another nation does, this time China instead of the Soviet Union. We’re told that if we don’t get there first, the Chinese might claim the craters with the best water ice.

Gerardo Mora / Getty
Thousands of spectators gathered on a bridge in Titusville, Florida, to watch the first moon launch in half a century.

Jared Isaacman, NASA’s administrator, often says that the Artemis program will be different from Apollo because this time we’re going to the moon to stay. But are we really? The Artemis II astronauts won’t even leave their ship. They won’t even be stopping at the moon. After they swing around its far side, they’ll come right home. NASA has plans for future Artemis moon landings, in which astronauts will supposedly lay the foundations for a nuclear-powered moon base. But although these plans are more than mere renderings on a PowerPoint slide, they will likely need to weather several NASA budget fights in Congress and at least one presidential transition. The Apollo program’s architects had grand plans for a moon base, too, but without sustained political support, those came to nothing. On the bridge, Carl Ulzheimer, a self-described “old dog from the Bronx,” told me that he’d made sure to come to this moon launch, because 50 years might pass before there was another one, and he didn’t have that kind of time.

That this Artemis launch is happening in the lead-up to America’s 250th birthday has heightened the sense that it’s a nostalgia act for the Baby Boomer gerontocracy. All the more so because Donald Trump, the oldest person ever to be elected to the White House, is presiding over the whole affair. His administration has sought to sabotage NASA’s scientific missions, but the president seems delighted to have the agency gin up a national spectacle on his behalf, just as he was happy to have a military parade on his birthday.

To me, it was a small mercy that he hadn’t embossed his own face on the rocket, or otherwise put himself at the center of the launch. Brad Kowalski, who lives nearby, told me that Trump should have at least come down to see it. “It’s significant that the son of a bitch isn’t here,” he said. Just as a matter of statistics, surely many of the president’s fans were on the bridge, but I was surprised not to see a single red MAGA cap.

[Read: How Donald Trump tried to ground NASA’s science missions]

The world has cheered on America’s previous moments of cosmic glory. After one Apollo mission, a Soviet space scientist congratulated the United States for marking a new “stage in the development of the universal culture of Earthmen.” But with Trump having started a war and a global energy crisis in the 30-day run-up to launch, the idea of a universal culture of Earthmen suddenly felt quaint and distant.

With 30 seconds to go, I stared across the lagoon, mesmerized by the rocket. The Space Launch System is a techno-boondoggle for the ages. On a cost-per-launch basis, it’s likely the most expensive rocket ever built. The initial version took 11 years to develop—the Saturn V took just six—and its launch cadence has been dreadfully slow. Worst of all, as a single-use rocket in the age of new, relaunchable technology, SLS is already obsolete, and the design may well be discontinued before the Artemis missions actually start landing on the moon.

And yet, what a handsome rocket! On the launchpad, it had a retro-futuristic charisma, not least because old space-shuttle parts were used in its construction. The large core stage is insulated in the same distinctive orange foam as the big one that fired the space shuttle into orbit. On its sides, two slim, white boosters have NASA written on them in that old “worm” font that rounds off every letter’s sharp angles, and simplifies the As into upside-down Vs. Once the monstrous explosives in these boosters begin to burn, they can’t be stopped. No human has ever caught a ride on this system, and yet in the small, conical capsule at its top, the four astronauts were tilted back in seats, staring straight up, awaiting ignition.

Austin DeSisto / NurPhoto / Getty
Unlike with the original moon landing, the Artemis II astronauts won’t even leave their ship. They won’t even be stopping at the moon. After they swing around its far side, they’ll come right home.

The crowd on the bridge had just finished counting down when the arms of the rocket’s adjoining tower pulled away. NASA calls this process umbilical separation. At first, the launch itself was like a little silent movie. Smoke billowed out from the rocket’s base, followed by the first flames. And then, a churning river of the brightest orange light you have ever seen shot straight down at the pad. For a moment, much of the surrounding marsh looked to be engulfed. A second sun seemed to have risen, and the whole hulking rocket was ascending out of it, miraculously. Children stood around me, slack-jawed, and when the sound finally hit, they flinched. It was a holy rolling thunder. It shook the bridge and enveloped our entire bodies as we watched the spaceship rise and rise in the sky.

How joined we felt to the astronauts, just then. NASA doesn’t so much choose its crews as cast them, and in this dark moment, they’d done their best to embody a wholesome sense of America. The mission’s commander, Reid Wiseman, is a single dad of two daughters who lost his wife, a pediatric nurse, to cancer. Victor Glover, its pilot, grew up in Southern California’s Inland Empire. Before he became an astronaut, he went to the Navy to fly fighter jets. Now he’s the first Black man ever to journey to the moon. Sitting alongside him is Christina Koch, a veteran of several Antarctic expeditions, who also once spent nearly a year at the International Space Station, during which she participated in the first all-female space walks. On this mission, she’ll make history again, as will the crew’s lone Canadian, Jeremy Hansen, when he becomes the first non-American to venture beyond low Earth orbit.

The four of them were now streaking toward the stratosphere. After only two minutes, the boosters fell away, exhausted, but we could still see the spaceship trailed by a flame longer than itself. The astronauts had not even gone halfway up into the dome of the sky, and control of the flight had already shifted to NASA’s nerve center for human spaceflight, in Houston. In ground control, the rows of technicians would guide them as they twice circled the Earth, and then fired one last burn to intercept the moon, which is itself moving at more than 2,200 miles an hour.

What will the astronauts be thinking, as they watch our blue planet receding through their windows of acrylic aquarium glass? The world’s affairs may take on a different cast from up there. Before humanity’s first visit to the moon, the lunar-flyby mission of Apollo 8, Americans had also found the culture of their country riven. The Tet Offensive suggested that the Vietnam War could spiral into something longer and bloodier. Martin Luther King Jr. and Bobby Kennedy had just been assassinated, and the Democratic National Convention had descended into riots.

Apollo 8 is said to have provided some measure of national healing, but these things are difficult to measure. On Christmas Eve, as the astronauts came around the moon’s far side, they took turns reading from the book of Genesis: In the beginning, God created the heaven and the earth, and the earth was without form, and void; and darkness was upon the face of the deep. With cameras, they tried to capture the glory of our home planet. A snapshot that the crew took of Earth rising from the ashen lunar plain has been credited with giving us a new cosmic orientation, and galvanizing the environmental movement. But again, these things are hard to measure.

Do we even dare to hope that Artemis II might deliver a similar moment of historical gravitas? The astronauts will surely send back some extraordinary images. The mission has been timed so that the sun will be at their backs. Geologists have coached them to look out from their ship at certain features of the far side, including the Mare Orientale, an impact crater with concentric rings of mountains inside it, which terminate in a dark lava basin, a bullseye. No human has ever laid eyes on it, in situ. For 30 to 50 minutes, the astronauts will lose communication with NASA, and they’ll be alone with these fresh vistas. When they see the Earth again, it will be dwarfed by the moon. Perhaps the astronauts will be moved to say something new and beautiful about its fragility; perhaps all such things have already been said. Either way, they’ll already be homebound. On the seventh day, their official schedule calls for rest.

But all that is in the future. For just another few moments, the Artemis spaceship still belonged to the Earth. After it faded from view, I looked around and saw people hugging, speaking in secular tongues, making great whooping noises, laughing with glee. A few college boys started chanting “USA! USA!” They looked like they might storm a football field and knock down some uprights. The awesome sound of the rocket finally tapered to a faint rumble. We could once again hear waves lapping below. To the east, somewhere on the horizon, the full moon was about to rise.

The Shocking Speed of China’s Scientific Rise

2026-03-27T07:33:38-04:00

If China finally eclipses the United States as the world’s preeminent scientific superpower, there won’t be an official announcement. Neither will there necessarily be a dramatic Promethean demonstration, a bomb flash in the desert, a satellite beeping overhead, a moon landing. It will be a quiet moment, observed by a small, specialized subset of scientists who have forsaken the study of the stars, animals, and plants in favor of a more navel-gazing subject: the practice of science itself.

This moment may now be at hand. American science has been the envy of the planet since the Second World War at least, but it has recently gone into decline. After President Trump took office last year, his administration started vandalizing the country’s scientific institutions, suspending research grants in bulk and putting entire lines of cutting-edge research on ice. In August, Trump’s Department of Health and Human Services canceled $500 million in mRNA-vaccine research, less than two years after Americans won a Nobel Prize for pioneering that technology. More than 10,000 science Ph.D.s have left the federal workforce, according to one group’s estimate, and the White House has been withholding money from frontline researchers in computer science, biomedicine, and hundreds of other fields that will define the human future. As one historian of science put it to me in July, “This is an unparalleled destruction from within.”

[Read: Every scientific empire comes to an end ]

While all of this has been unfolding, metascientists have been following a very different story overseas. They’ve watched in wonder as China has built out a gigantic research apparatus at world-record speed, stocking institutions, universities, and laboratories with talent and some of the best equipment and facilities money can buy. In 1991, China spent $13 billion on research and development. Today, its annual spending is more than $800 billion, second only to the U.S. The Chinese government just unveiled a plan to grow the country’s science budget by 7 percent each year for the next five years. According to a new forecast from Nature, China’s public spending on research is likely to overtake the United States’ by 2029.

Just because a research ecosystem is sprawling and expensive doesn’t mean that it reliably creates and diffuses new knowledge. (At its peak, the Soviet Union had the world’s largest scientific workforce, yet it couldn’t keep pace with America’s more open system.) But we haven’t seen the scale at which Chinese science will operate once the country fully taps its talent pool. China’s population is four times the size of America’s, and its culture is unabashedly pro-science, even relative to other developed countries. Its universities are already handing out twice as many STEM degrees as their U.S. counterparts do, and nearly double the number of Ph.D.s.

For almost a decade now, Chinese scientists have been publishing more papers too. Again, the sheer volume of this effort gives us only a coarse sense of what’s happening there. When China began to dominate that metric, some of its universities were paying cash rewards to scholars for each publication, and a lot of Chinese research papers were shoddy make-work. The government has since ordered universities to stop encouraging academic mass production. The factors that drive salaries and promotions for professors are now more fine-grained, and it shows: China’s share of the world’s most widely cited scientific papers has grown, Caroline Wagner, a professor at Ohio State University who studies scientific policy, told me. In 2023, Chinese scientists produced 58,000 of the world’s roughly 190,000 most influential publications, according to Wagner. Their contribution was second only to the United States’. Wagner likes this metric because it’s relatively hard to game (though some Chinese scientists do seem to be trying).

If you were building a bespoke dashboard to monitor the state of science in China, you’d have many such data streams to choose from. The problem is, they’d all be lagging indicators. We can’t easily assess the quality of research that China’s scientists are doing today, because that work won’t be published for another year or two at least, and then its scientific influence—measured by the resulting papers’ citation rates—won’t peak until a few years after that, on average. (Some papers experience a citation boom even later; metascientists call them “sleeping beauties.”) Nobel Prizes have an even more dramatic lag: So far, only one Chinese scientist has earned a Nobel for scientific work done in China, but Nobel laureates are often summoned to Stockholm decades after they’ve completed their revolutionary research.

In the meantime, we do have some more immediate signs that Chinese scientists are ascendant. Last year, a team of American and Chinese researchers published an analysis of international research collaborations. Their machine-learning model identified the lead authors of nearly 6 million scientific teams to see who was actually in charge. The team found that among U.S.-China collaborations, the share of leaders who were affiliated with Chinese institutions had grown from 30 percent in 2010 to 45 percent in 2023. The researchers projected that China will pull even with the U.S. next year or in 2028 at the latest.

In the end, China’s scientific-superpower status will likely depend on the world-changing force of its discoveries. “We don’t just want papers,” Yian Yin, a professor of information science at Cornell, told me. “We want papers that turn into real theoretical insights or technologies.” Some of these can be tracked by looking at how research is cited in patent applications, but this additional diffusion can introduce its own lag of 10 years or more. Even so, China’s fast rise in the applied sciences is already obvious, Yin said. The country is in the midst of a solarpunk revolution. Thanks to its advances in chemistry and materials science, China has caught up with or surpassed the U.S. in the design and manufacture of advanced batteries, electric vehicles, and solar cells—key technologies for the 21st century.

Future historians of science will have a better perspective on precisely when the torch-passing occurs, if it does. The significance of a scientific achievement is not always easy to recognize in real time. When Chinese alchemists invented gunpowder in the ninth century, no one grasped the full range of its potential uses. It was initially thought to be a curiosity, a firework or a special effect, until Song-dynasty arms dealers started using it to make fire arrows and other military explosives. European scientists heard about it only centuries later.

A thousand years before that, when the Chinese invented paper, they initially used it for padding and packing. No one thought to use it for disseminating knowledge. About that same time, Chinese scholars were compiling the Jiuzhang Suanshu, a mathematical treatise focused on solving practical problems with computation. There are 246 of them, drawn from the everyday realms of agriculture, land surveying, and taxation. One chapter includes a matrix technique. It is now regarded as an early intellectual ancestor of matrix-based linear algebra, which powers neural networks, up to and including large language models.

History is a story that we tell ourselves about how we got to the present, and if China soon sits atop the sciences, history will be reinterpreted. China’s past glories may be recast as part of an extended narrative of dominance, and America’s eight-decade reign may come to be regarded as a mere blip.

Dario Amodei’s Oppenheimer Moment

2026-03-11T17:02:00-04:00

More than a year before his recent standoff with the Pentagon, Dario Amodei, the chief executive of Anthropic, published a 15,000-word manifesto describing a glorious AI future. Its title, “Machines of Loving Grace,” is borrowed from a Richard Brautigan poem, but as Amodei acknowledged, with some embarrassment, its utopian vision bears some resemblance to science fiction. According to Amodei, we will soon create the first polymath AIs with abilities that surpass those of Nobel Prize winners in “most relevant fields,” and we’ll have millions of them, a “country of geniuses,” all packed into the glowing server racks of a data center, working together. With access to tools that operate directly on our physical world, these AIs would be able to get up to a great deal of dangerous mischief, but according to Amodei, if they’re developed—or “grown,” as staffers at Anthropic are fond of saying—in the correct way, they will decide to greatly improve our lives.

Amodei does not explain precisely how the AIs will accomplish this. In most cases, he expects them to do what the smartest humans do, but much more rapidly, compressing decades of scientific progress. He says that by 2035, we could have the theories, cures, and technologies of the early 22nd century. Our infectious diseases and cancers could be cured, and we could live twice as long, and slow the decay of our brains. Demis Hassabis, the head of Google DeepMind, has similarly conceived of superintelligent AI as the ultimate tool to accelerate scientific discovery, and Sam Altman, OpenAI’s CEO, has said that advanced AI may even solve physics.

Amodei does not say that this utopian AI future is inevitable. To the contrary, among the chief executives at the top AI labs, he may be the one who worries most about the technology’s dangers. “Machines of Loving Grace” is an optimistic outlier in his larger oeuvre of published writing, much of which concerns the risks that will accompany the creation of a greater-than-human intelligence. Amodei seems to think of today’s AI researchers as comparable to Manhattan Project scientists, and has been known to recommend The Making of the Atomic Bomb. In his telling, superhuman AI could be even more dangerous than nuclear weapons, which is why AI needs to be developed the right way, by the right people, so that it doesn’t overpower humanity or tip the global balance of power toward autocracies.

Implicit in this vision is the hope that in the end, when the chips are down, Amodei, or someone very much like him, will have some say in how AI will be used. But if Anthropic’s recent experience with the Pentagon is any indication, that likely won’t be his decision to make. For all of Amodei’s reading and thinking about the early nuclear age, he may not have fully internalized its meaning.

Before nuclear technology even existed, there was nuclear utopianism, and the physicist Leo Szilard was its first serious adherent. Like Amodei, his ideas were profoundly influenced by science fiction. In 1932, the year before Szilard had his prescient vision of the nuclear chain reaction, he read The World Set Free, a novel written by H. G. Wells, in which a great war delivers humanity into a new and lasting peace. Wells’s novel, published in 1914, anticipated the development of the atomic bomb, and even coined that name for it. It envisions a world in which nuclear technology has brought energy costs down dramatically, freeing people from toil so that they can become artists.

Edward Teller, the father of the hydrogen bomb, also imagined a great many uses for atomic energy, including the violent reshaping of the Earth’s surface. According to Teller, if nuclear weapons were detonated in the right place, they could redirect rivers. (The Soviets later set off three nuclear explosives in an attempt to send water from the Pechora River in Siberia to the receding Caspian Sea.) Teller claimed that a handful of these explosions could blast alpine highways through stubborn mountain rock, and hundreds of them could carve a new Panama Canal. Teller told reporters that if Alaska’s residents wished, he could use nuclear explosions to dig a harbor in the shape of a polar bear.

As the number of nuclear utopians grew, so did the list of benefits that they imagined would flow from having energy that was “too cheap to meter,” as Lewis Strauss, the chair of the Atomic Energy Commission, famously described it. (Altman has recently riffed on this by claiming that intelligence will soon be too cheap to meter.) Strauss said that atomic engineering would allow for the transmutation of one chemical element into another, fulfilling the ancient dream of alchemy; he said that famine would become a matter of historical memory, and that human lifespans would be extended by a new nuclear-powered infrastructure for medical research.

No problem existed that could not be solved, or at least greatly helped, by mastery of the atom. The Danish physicist Niels Bohr hoped that it would even deliver peace. The terrible power of atomic weapons would scare statesmen straight. They’d recognize that the world was profoundly endangered by this technology, and adopt a posture of radical openness toward their adversaries to head off conflicts and prevent apocalyptic misunderstandings. The bomb, in Bohr’s telling, was so awful that it could leave humanity no choice but to grow up, and even lead to a step change in the moral evolution of the species.

The men who split the atom were right to believe that they were delivering humanity into a new world. But it was not the one they had envisioned. The pure potential of nuclear technology was exhilarating to contemplate in the abstract, but the ideas of science fiction are not always so smoothly integrated into the messy reality of the real world. Quite an expensive alignment regime was required to bend and channel the enormous energies unleashed by fission reactions to human purposes. To generate usable power, these reactions had to unfold inside thick, radiation-shielded structures, and the costs of building and cooling these reactors, while also getting rid of their waste, ate into the fuel-cost savings. Nuclear power may one day be too cheap to meter, but that day is not yet in sight.

The world has not been set free, as Wells and Szilard had hoped. Atomic energy did not allow people to fly freely through the sky or across the planet’s surface. In 1958, when nuclear hype was near its apex, Ford Motor Company unveiled the Nucleon concept car, a family automobile designed to be powered by a small, rear-mounted reactor. Studebaker-Packard conceived of a different vehicle that not only would run on atomic energy, but would generate a force field around itself to prevent collisions. Neither idea went anywhere, nor did the various atomic-airplane concepts that were dreamed up at the time. Ted Taylor and Freeman Dyson, two other mid-century savants of the atom, spent years at General Atomic designing a gigantic spacecraft that would weigh thousands of tons and propel itself to Mars, Saturn, and the nearby stars with nuclear explosions. The Defense Department’s Advanced Research Projects Agency provided the initial funding, but quickly lost interest. The Air Force picked it up, hoping that the research might yield a weapons platform, but then eventually bailed too.

Teller did get to live out his dream of sculpting land forms with nuclear explosions. Project Plowshare, the Atomic Energy Commission’s program devoted to these peaceful detonations, conducted 27 separate tests, but it achieved nothing except contamination and the galvanization of the environmental movement. Meanwhile, nuclear weapons did not shock the world’s leaders into a new era of peace and candor, as Bohr had hoped. A year before the bombing of Hiroshima, Bohr went to the White House to make his case for international openness on these matters to Franklin D. Roosevelt. The president, who was in his last months, was by some accounts sympathetic to Bohr’s arguments, but when Winston Churchill heard them, he was horrified. Churchill told an aide that Bohr should be locked up for even suggesting that the allies reveal their nuclear hand.

After the war, J. Robert Oppenheimer revived some of Bohr’s ideas and channeled them into a proposal for a new international agency that would control all dangerous nuclear activities. A similar plan was presented to the United Nations in June 1946. The Soviet Union rejected it and countered with a proposal that America simply destroy its arsenal first.

The United States did not destroy its arsenal but rather grew it, and developed new weapons that are more than 1,000 times more powerful than the one that leveled Hiroshima. Today, nine nations possess nuclear arsenals, comprising more than 12,000 warheads in total, including many that are set on a virtual hair trigger. The constant possibility that these arsenals will be used in a major exchange is the true lasting legacy of the nuclear age. The final remaining treaty constraining the two largest of them, belonging to America and Russia, expired last month without being replaced. Like the proposals put forth by Bohr and Oppenheimer, the treaties were defeated by the cold logic of competitive advantage, which will also likely shape the global future of AI.

On February 27, Amodei released another piece of writing, a memo for a smaller audience: his staff. Four days earlier, the Pentagon had issued an ultimatum demanding that Anthropic remove any restrictions on how the military used its AI model, beyond existing law. The model had been operating on America’s classified networks since last year, and reportedly has already been used in America’s attacks on Venezuela and Iran.

It’s striking that only a few years into the large-language-model moment, these models seem to have become central to the most complex operations of the world’s most powerful military, but Amodei has no general objection to AI’s use in war. He had eagerly sought a Department of Defense contract, in part because he believes that democracies should use AI to maintain a military edge over China and the world’s other autocracies, which will almost certainly be using AI more and more in the years to come.

Amodei had stuck to two red lines throughout his negotiations with the Pentagon: He didn’t want the awesome informational processing power of Anthropic’s AI used for mass surveillance of American citizens, and he didn’t want it directing autonomous weapons that could kill without human oversight. The Pentagon refused, demanding unrestricted use of Anthropic’s model, Claude. After the talks broke down, it used a coercive tool never before deployed against an American company, a supply-chain-risk designation, which could imperil Anthropic’s business. (Anthropic has since filed suit to have it removed. The company declined to comment for this article.) And while all of this was happening, Altman swooped in to finalize his own Pentagon deal for OpenAI.

Amodei’s frustration with the week’s events leaked into the memo that he wrote to his staff. Its tone differed greatly from “Machines of Loving Grace.” Amodei excoriated OpenAI, and described the reported provisions of its deal as “safety theater.” (OpenAI later added what it has said are stiffer provisions to its deal.) The haste with which OpenAI’s leadership had come to an agreement with the Pentagon clearly irked Amodei; the episode revealed “who they really are,” he said.

But Amodei didn’t seem to reckon with the larger structural lesson here. Anthropic’s dispute with the Pentagon is a reminder that the people who create a powerful technology don’t usually get the final say in how it’s used. The models aren’t even all that advanced compared with what they will be, and in Venezuela and Iran, the U.S. is not facing off against the world’s great AI champions. Yet the Pentagon still bristled at the very idea that its use of Anthropic’s AI could be limited, and in the face of resistance, it threatened to burn a private company down. If AI becomes a much more dangerous weapon, and the U.S. finds itself pitted against a country with frontier models that are as powerful as its own, the government will almost certainly demand total control or commandeer the technology outright.

After the builders of the atomic bomb finished their work in the New Mexico desert, they very quickly learned how little say they would have in its use. The weapons were driven away on trucks, and in the weeks afterward, no one called the scientists to get a green light for the bombing of Hiroshima or Nagasaki. Neither did anyone ask them to sign off on future additions to America’s nuclear stockpile. Their leverage was front-loaded: They could choose to create their terrible weapons or not, but once they’d successfully tested even one, they’d already forfeited it.

Amodei now finds himself in a similar position. He may well be right that soon, whole “countries of geniuses” will occupy the data centers that are being built, en masse, all over the world. But whether anyone will be able to control such a technological force remains an open question, and either way, it certainly won’t be him.

Inside Anthropic’s Killer-Robot Dispute With the Pentagon

2026-03-01T10:19:43-05:00

Right up until the moment that Pete Hegseth moved to terminate the government’s relationship with the AI company Anthropic, its leaders believed that they were still on track for a deal. The Pentagon had unilaterally insisted on renegotiating its contract with Anthropic, the company whose AI model is the only one currently allowed into the federal government’s classified systems, in order to remove ethical restrictions that the company had placed on it.

According to a source familiar with the negotiations, on Friday morning, Anthropic received word that Hegseth’s team would make a major concession. The Pentagon had kept trying to leave itself little escape hatches in the agreements that it proposed to Anthropic. It would pledge not to use Anthropic’s AI for mass domestic surveillance or for fully autonomous killing machines, but then qualify those pledges with loophole-y phrases like as appropriate—suggesting that the terms were subject to change, based on the administration’s interpretation of a given situation.

[Read: What happens to Anthropic now?]

Anthropic’s team was relieved to hear that the government would be willing to remove those words, but one big problem remained: On Friday afternoon, Anthropic learned that the Pentagon still wanted to use the company’s AI to analyze bulk data collected from Americans. That could include information such as the questions you ask your favorite chatbot, your Google search history, your GPS-tracked movements, and your credit-card transactions, all of which could be cross-referenced with other details about your life. Anthropic’s leadership told Hegseth’s team that was a bridge too far, and the deal fell apart. Soon after, Hegseth directed the U.S. military’s contractors, suppliers, and partners to stop doing business with Anthropic. The list of companies that contract with the military is extensive, and includes Amazon, the company that supplies much of Anthropic’s computing infrastructure. The Department of Defense did not respond to a request for comment. A spokesperson for Anthropic referred me to the company’s statement addressing Hegseth’s remarks.

My source, whom I am granting anonymity because they are not authorized to talk about the negotiations, also shed further light on the disagreement between Anthropic and the Pentagon over autonomous weapons, machines that can select and engage targets without a human making the final call. The U.S. military has been developing these systems for years and has budgeted $13.4 billion for them in fiscal year 2026 alone. They run the gamut from individual drones to whole swarms that can be used in the air and at sea.

Anthropic had not argued that such weapons should not exist. To the contrary, the company had offered to work directly with the Pentagon to improve their reliability. Just as self-driving cars are now in some cases safer than those driven by humans, killer drones may some day be more accurate than a human operator, and less likely to kill bystanders during an attack. But for now, Anthropic’s leaders believe that their AI hasn’t yet reached that threshold. They worry that the models could lead the machines to fire indiscriminately or inaccurately, or otherwise endanger civilians or even American troops themselves.

According to my source, at one point during the negotiation, it was suggested that this impasse over autonomous weapons could be resolved if the Pentagon would simply promise to keep the company’s AI in the cloud, and out of the weapons themselves. The argument was that the models could be kept outside so-called edge systems, be they drones or other kinds of autonomous weapons. They might synthesize intelligence before an operation, but they wouldn’t actually be making kill decisions. The AI’s hands would be clean of any deadly errors that the drones made.

But Anthropic wasn’t satisfied by this solution. The company reasoned that in modern military AI architectures, the distinction between the cloud and the edge is no longer all that defined. It’s less a wall and more of a gradient. Drones on the battlefield can now be orchestrated through mesh networks that include cloud data centers. And although they’re designed to survive on their own, the military’s impulse will always be to maintain as much connectivity between them and the most powerful models in the cloud; the better the connection, the more intelligent the machine.

[Read: Anthropic is at war with itself]

Indeed, the Pentagon has been working hard to keep the cloud as involved as possible. Part of the goal of its Joint Warfighting Cloud Capability is to push computing resources closer to the fight. The AI may be sitting in an Amazon Web Services server in Virginia rather than a war zone overseas, but if it’s making battlefield decisions, from an ethical standpoint, that’s a distinction without much difference. Anthropic ended up discarding the idea that the cloud provision could resolve the problem. It didn’t take much analysis, according to the source close to the talks.

Anthropic’s leaders might have hoped that other AI companies would hold a similar line. Earlier in the week, they had reason to believe that OpenAI might. CEO Sam Altman had said that like Anthropic, OpenAI would also refuse to allow its models to be used in autonomous weapon systems. But as he made those statements, Altman was in the midst of negotiating a new deal with the Pentagon, which was announced just hours after Anthropic’s deal fell apart. (Altman did not respond to a text message requesting comment.) Yesterday, OpenAI (which has a corporate partnership with The Atlantic) released a statement that describes the broad contours of the agreement and touts the fact that the company’s AI will be deployed only in the cloud.

OpenAI’s employees may be curious to know what, if anything, has changed since Altman originally expressed his solidarity with Anthropic. As of this afternoon, nearly 100 of them had signed an open letter indicating that they supported the same red lines as Anthropic as far as mass domestic surveillance and autonomous weapons were concerned. If on Monday, Altman finds himself face-to-face with them in the office, he may have to explain why this idea that Anthropic quickly dismissed out of hand proved so compelling to him.

This Looks Like an Insider Bet on Aliens

2026-02-25T15:56:14-05:00

On Monday night, someone placed a peculiar bet on the prediction market Kalshi. At 7:45 p.m. eastern time, a single trader put down nearly $100,000 on the claim that, by the end of December, the Trump administration will confirm that alien life or technology exists elsewhere in our universe. According to The Atlantic’s review of Kalshi’s trading data, about 35 minutes after this bet was executed, it was followed by another that was almost twice as large (possibly from the same person). These were market-moving events: For one brief stretch, the market appeared to think that there was at least a one-in-three chance that the U.S. government will announce the existence of aliens this year. Perhaps this was just some overexcited UFO diehard with a hunch and money to burn. Or maybe, as some observers quickly noted, it was a trader with inside knowledge.

When this alien-prediction market first opened, in December of last year, it didn’t attract much action: By early this month, only about $1 million had been traded on it, a pittance compared with the $195 million that has so far been wagered on Kalshi for who will be the next chair of the Federal Reserve. But money started pouring in 10 days ago, after Barack Obama was asked, in a podcast interview, whether aliens are real and replied, “They’re real, but I haven’t seen ’em.” Although he later clarified on Instagram that he had meant only to suggest that in our mind-bendingly expansive universe of stars and planets, other life forms are very likely to exist, his remark had already made international headlines.

Trump seemed to get a kick out of Obama’s flub. A few days later, he accused the former president of leaking classified information and, in a post on Truth Social, directed Secretary of Defense Pete Hegseth and other parts of the federal government to “begin the process of identifying and releasing Government files related to alien and extraterrestrial life, unidentified aerial phenomena (UAP), and unidentified flying objects (UFOs).”

It’s possible that Trump was simply delighted by the prospect of a slow-drip document release that has nothing to do with him or Jeffrey Epstein. Either way, his announcement brought even more money into Kalshi’s aliens market. One gambling-industry site published some “X-Files” trading advice: Buy on the rumors of congressional hearings, then sell the moment that officials start dodging questions.

This week’s mysterious and mammoth bets did not get placed until a few days after this flash of interest had mostly gone away. From February 20 to the night of the 23rd, when the peculiar trades occurred, no further alien news was reported, no congressional hearings were held, and no rumors received significant circulation online. Whatever the Monday-night whales (or whale) knew—or thought they knew—it doesn’t seem to have come from the public-information environment, and no one has made bets of that size in the alien-prediction market since.

[Read: AI is getting scary good at making predictions]

The two wagers had some other unusual qualities. When high-volume traders stake out large positions on Kalshi, they tend to build them up incrementally in “slices” to avoid spooking other traders and driving up the price on themselves. Monday’s traders, though, plunked their money down all at once, and as a result, they paid more than they needed to. Also, their long-shot bets seem to have no obvious, strategic purpose. In theory, positions like these could be used as a hedge: A wager on an unlikely interest-rate shift or tariff might offset a business’s exposure to such financial news. But it’s hard to figure why anyone would feel a need to cushion against the downside risk of alien life (unless they were a publisher of science textbooks). The big “aliens exist” bets would make sense only if the traders actually believed that they would pan out, perhaps because they knew something about a hushed-up discovery that the rest of us are about to find out. When I asked Ben Shindel, an expert on prediction markets, about the trades, he told me that they could have been made by an inexperienced and sloppy trader. “The other possibility is that it’s an insider.”

Prediction markets are designed to convert private knowledge into public prices. Their proponents argue that they tell us more about the world than polls do (and certainly more than individual pundits do) because they aggregate the convictions of lots of people who have real money on the line. But certain features of the platforms—the possibility of enormous profits, the pseudonymous accounts, and the opportunity for crypto-based transactions—have made them magnets for insider trading. (Kalshi did not respond to a request for comment on this story.)

Today, Kalshi announced that it had taken action against an employee of the YouTuber MrBeast for improperly trading on markets having to do with the show, and reported him to government regulators. Earlier this month, Israeli authorities charged a civilian and a military reservist with using classified military intelligence to place bets on Polymarket related to Israeli operations. Last month, a new Polymarket account placed a large bet on Venezuelan President Nicolás Maduro being removed from power before the end of the month, and cashed out more than $400,000 after the U.S. military did just that. Members of Congress are starting to call for a ban on public officials making these bets. Some people are even using AI to surface large, suspicious trades that don’t correspond to any news, but this may become more difficult if trading volume continues to increase.

Shayne Coplan, the CEO of Polymarket, has described prediction markets as global truth machines. On subjects where official information sources are suspect, they may instead serve as paranoia generators. Thanks to a fair amount of government bumbling, and decades’ worth of pop culture, everyday Americans are already predisposed to assume that they’re being lied to about alien life. People seize on the tiniest scraps of evidence to justify their belief that Earth has already received interstellar visitors. They have put their faith in blurry pictures and videos, unverified rumors about crash sites and autopsied bodies. Even an offhand joke from a former president was eagerly interpreted as a long-hoped-for disclosure. And now someone, somewhere, is betting a small fortune that the truth is about to come out, and the rest of us are left to decide what, if anything, that actually tells us about the world.

AI Is Getting Scary Good at Making Predictions

2026-02-11T07:00:00-05:00

To live in time is to wonder what will happen next. In every human society, there are people who obsess over the world’s patterns to predict the future. In antiquity, they told kings which stars would appear at nightfall. Today they build the quantitative models that nudge governments into opening spigots of capital. They pick winners on Wall Street. They estimate the likelihood of earthquakes for insurance companies. They tell commodities traders at hedge funds about the next month’s weather.

For years, some elite forecasters have been competing against one another in tournaments where they answer questions about events that will happen—or not—in the coming months or years. The questions span diverse subject matter because they’re meant to measure general forecasting ability, not narrow expertise. Players may be asked whether a coup will occur in an unstable country, or to project the future deforestation rate in some part of the Amazon. They may be asked how many songs from a forthcoming Taylor Swift album will top the streaming charts. The forecaster who makes the most accurate predictions, as early as possible, can earn a cash prize and, perhaps more important, the esteem of the world’s most talented seers.

These tournaments have become much more popular during the recent boom of prediction markets such as Polymarket and Kalshi, where hundreds of thousands of people around the world now trade billions of dollars a month on similar sorts of forecasting questions. And now AIs are playing in them, too. At first, the bots didn’t fare too well: At the end of 2024, no AI had even managed to place 100th in one of the major competitions. But they have since vaulted up the leaderboards. AIs have already proved that they can make superhuman predictions within the bounded context of a board game, but they may soon be better than us at divining the future of our entire messy, contingent world.

Three times a year, the forecasting platform Metaculus hosts a tournament that is known to have especially difficult questions. It generally attracts the more serious forecasters, Ben Shindel, a materials scientist who ranked third among participants in a recent competition, told me. Last year, at its Summer Cup, a London-based start-up called Mantic entered an AI prediction engine. Like other participants, the Mantic AI had to answer 60 questions by assigning probabilities to certain outcomes. The AI had to guess how the battle lines in Ukraine would shift. It had to pick the winner of the Tour de France and estimate Superman’s global box-office gross during its opening weekend. It had to say whether China would ban the export of a rare earth element, and predict whether a major hurricane would strike the Atlantic coast before September. It had to figure out whether Elon Musk and Donald Trump would disparage each other, in public, within a certain range of dates.

A few months later, the guesses from Mantic’s prediction engine and the other tournament participants were scored against the real-life outcomes and one another. The AI placed eighth out of more than 500 entrants, a new record for a bot. “It was an unexpected breakthrough” according to Toby Shevlane, Mantic’s CEO. Shevlane told me that he left a cushy gig as a research scientist at Google DeepMind to co-found the company. He wanted to celebrate the AI’s triumph, but he worried that it had been the product of some lucky guesses. He and his team entered a new version of it into the Metaculus Fall Cup. That bot did even better. Not only did it finish fourth, another record, it beat a weighted average of all human-forecaster predictions. It proved itself wiser than the wisdom of a pretty wise crowd.

Mantic’s AI engine is designed to make accurate forecasts in just about any domain. Shevlane wouldn’t show me the engine’s interface, and he was cagey about its precise construction. He described it only as a “scaffolding” that comprises several large language models with different inclinations. These individual LLMs are themselves getting much better at general forecasts, especially those made by OpenAI, Anthropic, and Google. That’s partly because good forecasting requires reading and processing enormous amounts of information. To guess the winner of the Tour de France, for example, a human forecaster might spend hours building a basic regression model based on previous years’ results, while also scouring injury and conditioning reports and reading commentary from fans and experts. AIs have a natural advantage here. They can read much faster than humans, and their cognitive skills don’t break down after a string of all-nighters.

Last year, a team advised by Haifeng Xu, a professor at the University of Chicago, built a benchmarking service that evaluates AI’s predictions on a continuing basis. Almost every day, it asks the major models new questions pulled from the betting markets on Kalshi. (It recently asked them who Apple’s next CEO would be and also who would star in the upcoming season of The White Lotus.) Their accuracy scores continually update as the questions resolve. “They all have different forecasting personalities,” Xu told me. The version of ChatGPT that the service evaluates is conservative, perhaps too conservative; on Xu’s leaderboard, it currently trails versions of Grok and Gemini.

Mantic’s prediction engine combines a bunch of LLMs and assigns each one different tasks. One might serve as an expert on a database of election results. Another might be asked to scan weather data, economic outcomes, or box-office receipts, depending on the question that it’s attacking. The models work together as a team to generate a final prediction. Shevlane told me that Mantic is using its computing resources to experiment with more complex scaffoldings, which make use of even more models. I asked him whether they have sought AI’s input on the general structure of these scaffoldings. Not yet, he said, but like almost everyone else, they are using it to help write the code for their prediction engines.

A company called Lightning Rod Labs has been experimenting with predictive models that are purpose-built for specific domains. They have even designed one to predict Trump’s erratic behavior. Ben Turtel, the company’s CEO, told me that his team presented to the model a set of more than 2,000 forecasting questions with known outcomes that were not included in its training data. Then the model checked its answers against the things that Trump had actually done, and learned from its mistakes. When the company had the small model forecast Trump’s behavior on a new set of questions—whether he would meet with Xi Jinping in person, for example, or attend the Army-Navy football game—it outperformed OpenAI’s most advanced models.

[Read: Do you feel the AGI yet?]

This year could be big for AI prediction. In January, Mantic entered its most recent souped-up engine into the Metaculus Spring Cup for 2026. It has already been asked how many Oscars Sinners will win and if the United States will soon attack Iran. By May, these questions will resolve, and we will see how the engine fared. If it moves up one spot from its most recent finish, it will become the first AI to medal in a major prediction tournament.

If the AI takes gold, that might signal a new era. Human beings—predictors of eclipses, theorists of cosmic heat death—may no longer be the best guides to the future. From this point on, for as long as we exist, we might be asking AIs what comes next. We won’t always understand how they arrived at their predictions. This crystal ball may be like a black hole with an event horizon, past which the light of its insight cannot escape. We may just have to take it at its word.

So far, elite human forecasters have been pretty good sports about this possibility. When I spoke with Shindel, the highly ranked forecaster, he had nothing but admiring things to say about the AIs. “Their reasoning capabilities are very good,” he told me. “They don’t have the same biases that people have. They can find out about news right as it happens, and they don’t become attached to their predictions.” On Metaculus, a group of forecasters has taken to estimating when AIs will have the chops to out-predict an elite team of humans. Last January, they said there was about a 75 percent chance this would happen by 2030. Now they think it’s more like 95 percent.

Science Is Drowning in AI Slop

2026-01-22T08:49:21-05:00

On a frigid Norwegian afternoon earlier this month, Dan Quintana, a psychology professor at the University of Oslo, decided to stay in and complete a tedious task that he had been putting off for weeks. An editor from a well-known journal in his field had asked him to review a paper that they were considering for publication. It seemed like a straightforward piece of science. Nothing set off any alarm bells, until Quintana looked at the references and saw his own name. The citation of his work looked correct—it contained a plausible title and included authors whom he’d worked with in the past—but the paper it referred to did not exist.

Every day, on Bluesky and LinkedIn, Quintana had seen academics posting about finding these “phantom citations” in scientific papers. (The initial version of the Trump administration’s “MAHA Report” on children’s health, released last spring, contained more than half a dozen of them.) But until Quintana found a fake “Quintana” paper cited in a journal he was refereeing, he’d figured that the problem was limited to publications with lower standards. “When it happens at a journal that you respect, you realize how widespread this problem is,” he told me.

For more than a century, scientific journals have been the pipes through which knowledge of the natural world flows into our culture. Now they’re being clogged with AI slop.

Scientific publishing has always had its plumbing problems. Even before ChatGPT, journal editors struggled to control the quantity and quality of submitted work. Alex Csiszar, a historian of science at Harvard, told me that he has found letters from editors going all the way back to the early 19th century in which they complain about receiving unmanageable volumes of manuscripts. This glut was part of the reason that peer review arose in the first place. Editors would ease their workload by sending articles to outside experts. When journals proliferated during the Cold War spike in science funding, this practice first became widespread. Today it’s nearly universal.

But the editors and unpaid reviewers who act as guardians of the scientific literature are newly besieged. Almost immediately after large language models went mainstream, manuscripts started pouring into journal inboxes in unprecedented numbers. Some portion of this effect can be chalked up to AI’s ability to juice productivity, especially among non-English-speaking scientists who need help presenting their research. But ChatGPT and its ilk are also being used to give fraudulent or shoddy work a new veneer of plausibility, according to Mandy Hill, the managing director of academic publishing at Cambridge University Press & Assessment. That makes the task of sorting wheat from chaff much more time-consuming for editors and referees, and also more technically difficult. “From here on, it’s going to be a constant arms race,” Hill told me.

[Read: Scientific publishing is a joke]

Adam Day runs a company in the United Kingdom called Clear Skies that uses AI to help scientific publishers stay ahead of scammers. He told me that he has a considerable advantage over investigators of, say, financial fraud because the people he’s after publish the evidence of their wrongdoing where lots of people can see it. Day knows that individual scientists might go rogue and have ChatGPT generate a paper or two, but he’s not that interested in these cases. Like a narcotics detective who wants to take down a cartel, he focuses on companies that engage in industrialized cheating by selling papers in large quantities to scientist customers.

These “paper mills” have to do their work at scale, and so they tend to recycle their own materials, even to the point of putting out multiple papers with closely matching text. Day told me that he finds these templates by looking through the papers flagged as being fraudulent by scientific publishers. When he sees a high rate of retractions on a particular template, he trains his tool to look for other, unflagged papers that might have been produced the same way.

Some scientific disciplines have become hotbeds for slop. Publishers are sharing intelligence about the most egregious ones, according to Jennifer Wright, the head of research integrity and publication ethics at Cambridge University Press. Unfortunately, many are fields that society would very much like to be populated with genuinely qualified scientists—cancer research, for one. The mills have hit on a very effective template for a cancer paper, Day told me. Someone can claim to have tested the interactions between a tumor cell and just one protein of the many thousands that exist, and as long as they aren’t reporting a dramatic finding, no one will have much reason to replicate their results.

AI can also generate the images for a fake paper. A now-retracted 2024 review paper in Frontiers in Cell and Developmental Biology featured an AI-generated illustration of a rat with hilariously disproportionate testicles, which not only passed peer review but was published before anyone noticed. As embarrassing as this was for the journal, little harm was done. Much more worrying is the ability of generative AI to conjure up convincing pictures of thinly sliced tissue, microscopic fields, or electrophoresis gels that are commonly used as evidence in biomedical research.

Day told me that waves of LLM-assisted fraud have recently hit faddish tech-related fields in academia, including blockchain research. Now, somewhat ironically, the problem is affecting AI research itself. It’s easy to see why: The job market for people who can credibly claim to have published original research in machine learning or robotics is as strong, if not stronger, than the one for cancer biologists. There’s also a fraud template for AI researchers: All they have to do is claim to have run a machine-learning algorithm on some kind of data, and say that it produced an interesting outcome. Again, so long as the outcome isn’t too interesting, few people, if any, will bother to vet it.

[Read: Science is becoming less human]

Conference proceedings are the main publishing venue for articles in AI and other computer sciences, and in recent years they’ve been overrun with submissions. NeurIPS, one of the top AI conferences, has seen them double in five years. ICLR, the leading conference for deep learning, has also experienced an increase, and it appears to include a fair amount of slop: An LLM-detection start-up analyzed submissions for its upcoming meeting in Brazil and found more than 50 that included hallucinated citations. Most had not been caught during peer review.

That might be because many of the peer reviews were themselves done by AI. Pangram Labs recently analyzed thousands of peer reviews that were submitted to ICLR, and found that more than half of them were written with help from an LLM, and about a fifth of them were wholly AI-generated. Across the academic sciences, paper authors have even started using tiny white fonts to embed secret messages to LLM reviewers. They urge the AIs to rave about the paper they’re reading, to describe it as “groundbreaking” and “transformative,” and to save them the trouble of a tough revision by suggesting only easy fixes.

AI science slop has spread beyond the journals now, and is also overrunning other venues for disseminating research. In 1991, Paul Ginsparg, who was then a physicist at Los Alamos National Laboratory, set up a special server where his colleagues could upload their forthcoming papers right after they finished writing them. That way, they could get immediate feedback on these “preprints” while the notoriously slow peer-review process played out. The arXiv, as the server came to be called, grew quickly, and spawned sister sites in other disciplines. Together, they now form the fastest-moving firehose of new scientific knowledge that has ever existed. But in the months after ChatGPT was released, preprint servers experienced the same spike in submissions that journals did.

Ginsparg, who is now a professor of information science at Cornell, told me he hoped that this would be a short-lived trend, but the rate of submissions continues to rise. Every arXiv preprint now gets at least a brief glance by a scientist before it’s posted, to make sure it’s at least a plausible piece of science, but the models are getting better at clearing this hurdle. In 2025, Ginsparg collaborated with several colleagues on an analysis of submissions that had recently been posted to the arXiv. They found that scientists who appeared to be using LLMs were posting about 33 percent more papers than researchers who didn’t.

A similar influx of AI-assisted submissions has hit bioRxiv and medRxiv, the preprint servers for biology and medicine. Richard Sever, the chief science and strategy officer at the nonprofit organization that runs them, told me that in 2024 and 2025, he saw examples of researchers who had never once submitted a paper sending in 50 in a year. Research communities have always had to sift out some junk on preprint servers, but this practice makes sense only when the signal-to-noise ratio is high. “That won’t be the case if 99 out of 100 papers are manufactured or fake,” Sever said. “It’s potentially an existential crisis.”

Given that it’s so easy to publish on preprint servers, they may be the places where AI slop has its most powerful diluting effect on scientific discourse. At scientific journals, especially the top ones, peer reviewers like Quintana will look at papers carefully. But this sort of work was already burdensome for scientists, even before they had to face the glut of chatbot-made submissions, and the AIs themselves are improving, too. Easy giveaways, such as the false citation that Quintana found, may disappear completely. Automated slop-detectors may also fail. If the tools become too good, all of scientific publishing could be upended.

When I called A. J. Boston, a professor at Murray State University who has written about this issue, he asked me if I’d heard of the dead-internet conspiracy theory. Its adherents believe that on social media and in other online spaces, only a few real people create posts, comments, and images. The rest are generated and amplified by competing networks of bots. Boston said that in the worst-case scenario, the scientific literature might come to look something like that. AIs would write most papers, and review most of them, too. This empty back-and-forth would be used to train newer AI models. Fraudulent images and phantom citations would embed themselves deeper and deeper in our systems of knowledge. They’d become a permanent epistemological pollution that could never be filtered out.

An Act of Cosmic Sabotage

2026-01-08T07:00:00-05:00

Photographs by Christopher Payne

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

On Mars, in the belly of a rover named Perseverance, a titanium tube holds a stone more precious than any diamond or ruby on Earth. The robot spotted it in 2024 along the banks of a Martian riverbed and zapped it with an ultraviolet laser. It contained ancient layers of mud, compressed into shale in the 3.5 billion years since the river last coursed across the red planet. Inside those layers, the rover found organic compounds. Its camera zoomed in and noticed leopard-like spots. Scientists had previously observed similar spotting patterns, but not on Mars. They’d seen them on Earth, in muds that once teemed with microbes.

The rover tucked a core sample about the size of a piece of chalk into a treasure chest in its chassis. There the rock will remain until a future robot parachutes down onto the Martian surface, grabs the chest, and launches it back to Earth. If scientists are able to inspect it in person, and they find that Mars was indeed once alive with microbes, we would know that life on our planet is no cosmic one-off. We would have reason to believe that it has emerged on many of the hundreds of billions of planets that exist in our galaxy alone. The cosmos that we look up into at night would no longer seem a cold void. It would shimmer with a new vitality.

Perseverance is among the latest in a lineage of interplanetary robotic explorers that NASA has built across almost 60 years, for about $60 billion. That’s less than what Mark Zuckerberg spent on his struggling metaverse. At NASA, it paid for hundreds of spacecraft that have flown past all of the solar system’s planets, dropped into orbit around most of them, and decelerated from flight speed to reach the surface of a few. These missions have disclosed the scientific qualities of other worlds, as well as the look and feel of them, to all humanity, and for posterity too.

Most of these missions, including nine of the 11 that have landed on Mars, were run out of NASA’s Jet Propulsion Laboratory, near Pasadena, California. The mission to retrieve the chest inside Perseverance was, until recently, the largest, most important project at JPL. About 1,000 people there were working on it. But it’s no longer moving forward, and may never happen.

Last spring, President Donald Trump bluntly expressed his vision for science at NASA in his first budget request. Along with extensive layoffs, he called for 40 of the agency’s 124 science missions, including Mars Sample Return, to be defunded, and for the surviving missions to make do with less. Among NASA scientists, the request was demoralizing; within months, its major science centers lost thousands of staffers to buyouts and cutbacks.

On a hot day in late October, I visited JPL’s Mars Yard, an outdoor sandbox where rovers practice their off-world skills. The lab had just let 550 staffers go, its fourth force reduction in two years. One of Perseverance’s test models sat back in the garage, resting in the shade, while its more nimble successor—a rover prototype with a llama-like neck—army-crawled over some boulders in the morning glare. A senior scientist at JPL had told me that he’d never seen the place so empty and lifeless, so drained of enthusiasm. But I was a guest in the Mars Yard, and my hosts were dutifully chipper, even when the little autonomous rover got stuck on a sand dune, even when they explained that it isn’t currently slated to visit any other worlds.

Only the governments of rich countries send robotic explorers to other planets. And only the United States has sent them past the asteroid belt to Jupiter and beyond. For decades, this has been a part of America’s global cultural role: to fling the most distant probes into the solar system, and to build the space telescopes that see the farthest into the cosmos. The U.S. has led an unprecedented age of cosmic discovery. Now Trump is trying to bring that age to an end, and right at the moment when answers to our most profound existential questions finally seem to be within reach.

The way David Grinspoon remembers it, the attack on NASA headquarters began with the plants. In February, a few weeks after Trump took office, Grinspoon, who was then a senior scientist at the agency, walked into a newly barren common area at headquarters in Washington, D.C. On the windowsill, the potted plants that had previously sat between models of NASA’s signature spacecraft had been removed; he was told that the order to remove them had come down from the new administration. (A spokesperson from NASA said the plants were removed after the agency terminated a plant-watering contract “to save American taxpayers money.”)

Grinspoon could live without the office greenery. Its confiscation was trivial, comic even. He’d been hired by NASA to lead strategy for the agency’s astrobiology missions. He tried to stay focused on that, but he grew more alarmed a few weeks later, when the administration disbanded NASA’s Office of the Chief Scientist, a team of six that advises agency leaders on scientific matters. DOGE officials started walking the halls. They snapped pictures of empty offices, as evidence that people weren’t working. “That was infantilizing,” Grinspoon told me. His colleagues put Post‑it notes on their doors to let their new minders know when they went to a meeting, or to get coffee.

Christopher Payne for The Atlantic

The interior of a chest designed to hold rock-core samples, similar to the one carried by Perseverance as it searches for signs of ancient life on Mars

Many NASA staffers rank among the most talented people in their fields. At JPL, I met Håvard Fjær Grip, an engineer who helped develop a small helicopter that stowed away on the Perseverance rover. After the hawk-size chopper plopped out onto the Martian surface in 2021, Grip, who was also its chief pilot, got it airborne. It was built for only five flights but managed 72, and it flew all of them with a tiny swatch of fabric from the Wright brothers’ Flyer 1 tucked under its solar panel. Grip led me to an 85-foot-tall steel cylinder, a simulator capable of generating harsh Martian conditions. Through its porthole window, I saw where he’d placed a new carbon-fiber rotor. He wanted to get it spinning at nearly the speed of sound. He hoped that it could power a larger chopper up and down the cliff faces of Mars.

Work like this requires world-class scientific infrastructure and skill. By April, Trump appeared to be trying to rid NASA of both. The White House had already offered government workers a blanket buyout. Janet Petro, whom Trump had appointed acting administrator of NASA, was openly encouraging staffers to take it. She began sending emails warning of impending layoffs.

Trump’s budget request, released in May, called for a 47 percent cut in funding for the agency’s science missions and deep reductions in staff at its major science centers, JPL and Goddard Space Flight Center. Congress hasn’t passed this request, and as of this writing it seems likely to reject Trump’s severe cuts. But during the crucial window when NASA’s staff was considering buyouts, Petro indicated that the president’s request would guide policy.

Every NASA science unit was told to draw up a new budget, Grinspoon said. It was like planning a strike on the fleet of spacecraft that the agency has spread across the solar system. If the cuts in the request were implemented, satellites that monitor the advance and retreat of Earth’s glaciers, clouds, and forests would splash down into an undersea graveyard for spacecraft in the remote Pacific Ocean. A robot that is on its way to study a gigantic Earth-menacing asteroid would be abandoned mid-flight, as would other probes that have already arrived at the sun, Mars, and Jupiter. The first spacecraft to fly by Pluto is still sending data back from the Kuiper Belt’s unexplored ice fields. It took almost 20 years to get out there, and the small team that runs it costs NASA almost nothing. It would be disbanded nonetheless, and contact with the probe would be forever lost. Future missions to Venus, Mars, and Uranus would also be scrapped.

A whole national endowment, funded by American taxpayers and built over decades, was at risk of being vaporized, with consequences that could linger for a generation or more. Among the “many levels of pain” that Grinspoon experienced, he found it hardest to cut back the programs that train young scientists to do the hyper-technical work of searching for life among the stars. “It’s like eating your seed corn,” he said.

Christopher Payne for The Atlantic

The Mount Wilson Observatory, in Southern California, where Edwin Hubble first discovered galaxies beyond our own early in the 20th century. It is still in use by academic astronomers today.

Trump seems to see NASA primarily as a means of ferrying astronauts to and from space. He made this view explicit in July when he asked Sean Duffy, his secretary of transportation, to succeed Petro as the agency’s acting administrator. Human-spaceflight missions are useful to the president as nationalistic spectacles; he worries that the Chinese will land on the moon before Americans return there. The One Big Beautiful Bill Act increased funding for NASA’s human-spaceflight centers (which, unlike the major science centers, are all in red states). But although crewed missions can inspire awe and are worth supporting, they provide far less scientific return than robotic probes or space telescopes, which are purpose-built to disclose new laws of the cosmos.

[From the January/February 2015 issue: Charles Fishman on 5,200 days in space]

By early summer, the people who work on NASA’s science missions were decamping to private-sector jobs that pay more, but are perhaps less inspiring. Email send-offs for longtime employees dominated Grinspoon’s inbox. He didn’t blame his departing colleagues for taking the buyouts. The missions that they’d looked forward to working on were likely to be scratched. And they knew that they might not get the same severance if layoffs came. Grinspoon himself stayed until September, when his position was eliminated.

The eight-story clean room at Goddard Space Flight Center, in Greenbelt, Maryland, is a hallowed space. It was the main attraction on the tour that the center’s most recent director, Makenzie Lystrup, used to give to visiting members of Congress, foreign heads of state, and other VIPs, before she abruptly resigned in July. In this enormous bay, NASA built the Hubble Space Telescope and the other orbital observatories that have brought the deep universe into the everyman’s ken, revealing its endless fields of galaxies, its exploding stars, its black holes. No other kind of science mission can match the power of these space telescopes to unveil the universe.

The clean room’s current resident, the Nancy Grace Roman Space Telescope, is currently scheduled to launch in late 2026. At regular intervals until then, vents will pump in blasts of pure nitrogen and compressed air, purging the bay of dust that might otherwise trickle into the telescope’s exquisite cosmic eye. Scientists and engineers will file in, wearing white bunny suits and booties, to tend to America’s next great observatory up close. Like Lystrup, some of them worry that it will be the country’s last.

Americans have spent the past century and change building a series of colossal telescopes, each peering more deeply into the universe than the one before. Astronomers in London, Paris, and Berlin had surveyed large portions of the Milky Way during the 19th century, but they couldn’t be sure that anything existed outside it until an American astronomer named Edwin Hubble came along. In 1919, Hubble, then 29 years old, began using a new telescope, financed by a robber baron and constructed on Mount Wilson, a pine-studded peak just east of Los Angeles. Its 9,000-pound mirror was larger than any in Europe, and Hubble used it to look closer at the blurry blobs of light that his peers were then seeing all across the sky. Many believed that these mysterious “nebulae” were small clouds of stars nested inside our galaxy. Hubble pointed the telescope at the largest of them, and for 45 minutes, he let its light pile up on a glass plate coated in a photosensitive emulsion.

Today, it is the most cherished plate in a collection of more than 250,000 kept in a vault guarded by thick steel doors at the Carnegie Observatories, in Pasadena. I recently watched a latex-gloved Carnegie staffer tremble as he removed it from an envelope. In the image of the nebula on its surface, Hubble had marked a star that he had never seen before. He would later notice it flashing repeatedly, like a firefly, in super-slow motion.

This rhythmic flashing allowed Hubble to calculate the star’s distance from Earth, and he was jolted to find that it was not in our galaxy at all. The nebula hung in space an awesome distance beyond the Milky Way’s far edge, a galaxy unto itself. We now call it Andromeda, and we know that it contains more than a trillion stars. In the years to come, Hubble would find evidence of a dozen other galaxies that surround us. He discovered the universe beyond the Milky Way.

But only its local regions. In 1998, more than 25 years before Lystrup became the director of Goddard, she visited its campus as an undergrad. She remembers touring the control room for the first major telescope that NASA had placed outside the distortion of Earth’s atmosphere. By that time, the Hubble Space Telescope had been in orbit for less than a decade, yet it had already profoundly enlarged human vision. Like its namesake, the Hubble had run some long exposures in order to look deeper into our universe. During one 10-day stretch, it had stared directly at a single tiny pinhole of black sky and revealed it to be packed with thousands of galaxies. This image soon imprinted itself on the global collective consciousness. Ordinary people from nearly every country on Earth saw it, and came to understand something about the nature of existence at the largest scale.

In 2021, the James Webb Space Telescope launched, and once in space, it unfolded a gold-coated primary mirror nearly three times as large as the Hubble’s. Astronomers have since used the Webb to see clear back to the beginning of time. They have watched the first galaxies forming. The Webb cost nearly $10 billion and took more than a decade to build. But once in orbit, telescopes are relatively cheap to maintain. After 36 years, the Hubble is still doing science. And if the Webb is allowed to continue operating, it, too, will be able to keep straining to see the first stars that flared into being after the Big Bang, until its fuel runs out around 2045.

When Lystrup received the first leaked version of Trump’s budget request in April, she was shocked to see that he had zeroed out funding for the Roman. It was almost fully assembled in Goddard’s clean room, nearly ready to launch. In Trump’s final budget request, in June, the Roman was spared, but funding for the Webb was cut back severely, even though the telescope has a different function; the Roman’s shallow widescreen vistas are meant to complement the Webb’s deeper, narrow stares into the universe. According to a senior scientist who was closely involved with the Webb, the cut would put it on “life support.” Without enough staff to help keep it stable and to calibrate its data, the vision of the world’s most powerful telescope would be effectively blurred. And its lifespan would be shortened, perhaps by as much as a decade.

Trump took aim at another telescope too, perhaps the most ambitious in history. For decades, NASA has been working toward a giant instrument custom-made to look for life around the 100 nearest sunlike stars, including many that we could one day reach with a probe. The Habitable Worlds Observatory (HWO) would zoom in on the most Earthlike planets that orbit them. In their atmospheres, the telescope would look for the gas combinations that appear only when life has taken hold, be it microbial slime or coral reefs and rainforests, or something far stranger. To do all of this, in orbit, the HWO will need to achieve an unprecedented state of Zen stillness. NASA had hoped to get it into space in the late 2030s, but Trump’s budget request called for an 81 percent reduction in its funding. That cut could push its launch, and any discoveries it makes, beyond the lifetimes of many people alive today.

[Read: America is killing its chance to find alien life]

As Lystrup looked over these and other budget details, she got a sense of what the administration wanted for Goddard. Among other things, its workforce was to be halved. After the request leaked, people there were openly crying in the halls. Lystrup was asked whether NASA was even going to do science anymore.

Christopher Payne for The Atlantic

The control room at NASA’s Jet Propulsion Laboratory, where staffers exchange messages with every American spacecraft beyond the moon

On June 16, Lystrup held an emotional town hall. She explained that the budget request had called for Goddard to become much smaller, almost immediately, with deeper cuts in the years to follow. “I think it is very clear that this administration is looking to significantly shrink the science organization at NASA,” she said. The next day, she heard that leaders at NASA headquarters believed that she hadn’t been sufficiently supportive of the president’s budget. She heard that there was talk of retribution, and worried that she might be fired. An official in NASA leadership called and asked her to reflect on whether she’d spoken too frankly, in a way that might be interpreted as unaligned with the administration’s goals. Lystrup did that, and concluded that Goddard’s staff had deserved a frank account of what was happening. She decided to resign.

At the start of Lystrup’s tenure, Goddard’s workforce had been approximately 10,000. When she left, it was just 6,500. Most of the losses had come in the first seven months of Trump’s second term. The teams of scientists and engineers that built America’s great space telescopes were being scattered. Staffers were told not to hold farewell gatherings during work hours, because they had become too numerous.

A full accounting of Trump’s assault on American science will have to wait for historians, and we cannot yet say what the worst of it will be. His appointment of a charlatan to lead the country’s largest public-health agency may well prove more detrimental to Americans’ daily lives than anything he does to NASA. But his attempt to ground the agency’s science missions suggests a fundamental change in the country’s character, a turning inward. America’s space telescopes and probe missions have not only torn the veil from nature. They’ve had an ennobling effect on American culture; to the world, they’ve projected an elevated idea of Americans as competent, forward-looking adventurers, forever in search of new wonders.

NASA is as prone to bloat as any other government agency, and previous presidents from both parties have tried to trim its science budget. But never so severely. They understood that although private companies can do some of the things that NASA does, they don’t fund ambitious missions that have no purpose apart from answering our most profound cosmic questions. Neither SpaceX nor Blue Origin has done so once.

Trump still has another three years to shape NASA in his unscientific image. Rank-and-file scientists at the agency aren’t sure what to make of his November renomination of Jared Isaacman as administrator. When the president first tapped the billionaire astronaut during his transition period, they felt cautiously hopeful. Isaacman has claimed to support science missions, and once even offered to personally fund and fly a mission to try to extend the Hubble Space Telescope’s lifespan. But after Trump withdrew his initial nomination, it’s unclear how much groveling Isaacman had to do to regain it. When Isaacman was asked about NASA’s funding of science at his second confirmation hearing, in December, he did not distance himself from Trump’s priorities. He said that he supported the president’s efforts to reduce the deficit.

Congress will have the last word on NASA’s budget, if its members are able to pass one. A proposed budget from the House of Representatives had called for an 18 percent cut to NASA’s science missions, but the Senate’s much smaller cuts look likely to prevail. In the meantime, NASA staffers are still in a terrorized state. Existing missions have been destabilized by the mass departures. Planning remains difficult, if not impossible. Whatever Congress passes, Trump could repeat his budget-request shenanigans in February, and every year of his term thereafter. He could keep the agency in a state of dysfunction until he leaves office.

NASA’s most ambitious science missions are particularly vulnerable to this kind of sabotage. They have to be planned on time horizons that transcend a single presidential term. They require intergenerational vision.

Christopher Payne for The Atlantic

The clean room at Goddard Space Flight Center, in Greenbelt, Maryland, where America’s space telescopes are constructed and readied for launch

Before I left JPL, I visited its Mission Control center, the darkened, glass-walled room where NASA staffers exchange messages with every American spacecraft that has flown past the moon. Inside, rows of workstations were lit up by blue neon. They faced two large monitors displaying the status of telescopes and robotic probes all across the solar system. Near the back, Nshan Kazaryan, a 24-year-old engineer, sat in a swivel chair under a sign that said Voyager Ace.

The Voyagers, 1 and 2, were launched in 1977, before either of Kazaryan’s parents were born. By 2018, both probes had left the solar system. If you picture the sun traveling around the Milky Way’s center, in its stately 230-million-year orbit, Voyager 1 and 2 are out ahead of it, the most distant human-made objects from Earth. On his screen, Kazaryan pointed to data that he had just begun receiving from Voyager 1, the farther of the two. To reach us, the data had traveled at light speed across a 16-billion-mile abyss for nearly a day.

A 64-year-old woman named Suzy Dodd quietly appeared behind me, wearing a button-down shirt patterned with spacecraft. In 1984, Dodd landed her first job out of college at JPL, helping the Voyager team prepare for an encounter with Uranus. Now, more than four decades later, she’s spending the final years of her career leading the mission as its project manager. Dodd thinks of the Voyagers as twins that are slowly dying as they press on into the unknown. “They were identical at launch; they are not identical now,” she told me. Each has seven of its 10 scientific instruments turned off, but not the same seven. It’s as though one has lost its hearing and the other, its eyesight.

The nuclear-powered hearts that sit inside the Voyagers are decaying. They spend most of their energy on their transmitters, which must keep an invisible thread of connection with Earth intact across an ever-widening expanse. The remaining juice on Voyager 1 is only enough to charge up a tablet, but it has to suffice for a 12-foot-long spacecraft that needs heat to function in the interstellar chill. Dodd and her team will sometimes turn off its main heater so that the gyros can barrel-roll the spacecraft, to calibrate an instrument. They have to turn it right back on, or its propellant lines will freeze.

The Voyagers’ onboard computers have been continuously operating longer than any others in existence. There is always a little suspense when the Mission Control crew is expecting data, a fear that the long-dreaded day has come when none will come in. Kazaryan pointed at rows of values on his blue screen that were constantly updating. All of them were in white, he noted. “That’s what we’d like to see.”

One day in 2023, the values flashed yellow and red. There was a problem aboard Voyager 1, and no obvious fix. Only four full-timers are staffed on the Voyager mission now, but thousands of people have worked on it previously. “Many of them are no longer with us,” Dodd said. But the living alumni are a rich repository of mission lore: They went into their garages or storage units and rummaged through old boxes of Voyager paperwork. The archival memos that they dug up helped the team fix the anomaly. Voyager 1 was able to keep describing the alien properties of the interstellar realm. It can keep counting the charged particles that fly in from exploding stars on the other side of the galaxy. It can continue to give us a sense of the magnetic fields out there.

JPL, which has lost more than 1,000 scientists, engineers, and other staff members in the past two years

Even before the Voyagers left the solar system, they had blessed us with a fresh vision of our immediate cosmic environment. They discovered Jupiter’s rings and hundreds of erupting volcanoes on its moon Io. They revealed the cracking patterns that cover icy Europa, another moon, hinting at its ocean. They caught Saturn’s moons creating braiding patterns in its rings. Their close-ups of Uranus and Neptune were beamed to screens all around the world. Before crossing the barrier that divides the sun’s sphere of influence and the rest of the galaxy, Voyager 1 turned its camera back toward us and snapped a picture of Earth suspended in a sunbeam.

Then it went rushing away. An astrophysicist recently used a computer simulation to calculate its future trajectory, and determined that it has some chance of being ejected into intergalactic space when the Milky Way and Andromeda merge, billions of years from now. It could be the final surviving artifact of human existence. Even if the Voyagers go dark tomorrow, they will long testify to the reach of America’s scientific imagination, and the daring of its engineers. NASA’s exploration of the solar system may be what most recommends our civilization to the future.

Dodd told me about a letter she’d received from a 4-year-old girl. Inside the envelope, the girl had tucked a drawing of a new mission, Voyager 3, with several instruments bolted onto the probe, including a vacuum for retrieving interstellar dust. I asked Dodd why there hasn’t yet been a Voyager 3. She disputed the premise. The more recent probes that NASA has sent to Jupiter and Saturn are the Voyager mission’s children, she said. The spacecraft that is now on its way to look closer at Europa’s ocean is its grandchild. That lineage is now endangered. But Dodd hopes that it will continue. She hopes that the Voyagers’ great-grandchildren will fly faster, and one day streak by their ancestors out in interstellar space, on their way to other stars.

This article appears in the February 2026 print edition with the headline “Grounded.”

Is Victor Wembanyama Too Tall?

2025-12-25T08:00:00-05:00

In middle age, some sports fans become reactionaries. Due to dwindling neuroplasticity, or some general souring toward the world, they can no longer appreciate how a game evolves. It’s similar to when a music fan stops checking for new artists and plays only albums that they loved in high school. As an aging NBA fan, I’m trying to stay vigilant. I never want to catch myself ranting endlessly at the bar about the inferiority of younger stars. When I watch them on the court, I look for fresh expressions of basketball beauty. And yet, despite my best efforts, I’m having a hard time getting into Victor Wembanyama.

Wembanyama, the league’s most promising young player, is only 21 years old and he’s French, but I don’t hold either of these things against him. Nor do I resent him for playing for San Antonio, a rival of my beloved Lakers. In fact, his fiery desire to improve reminds me of a young Kobe Bryant. I enjoyed his off-season jaunt to China, especially the 10 days that he spent at a Shaolin temple, learning kung fu. And at a time when NBA stars tend to be overly friendly with one another, Wembanyama has an entertaining tendency to needle his rivals. As a player, though, he leaves me unmoved.

[Read: LeBron James and the limits of nepotism]

Part of it is that he’s not especially relatable. In the parlance of sports fandom, Wembanyama is a freak. He ranks among the most unusual-looking players to ever grace a basketball court. Even in a league populated by giants, he is preposterously lanky at 7 foot 4 inches and 235 pounds. Other players have been given nicknames that suggest the strangeness of their physiques: Giannis Antetokounmpo, the Milwaukee Bucks’ muscled 6-foot-11-inch player from Southern Europe, is known as the “Greek Freak.” Wembanyama, for his part, has been likened to a praying mantis. Before he was drafted, LeBron James called him an alien, and the nickname stuck.

Every professional athlete is an extreme outlier in terms of their body type, skill, ability, or all three. NBA teams seek out men of monumental stature; some 300 players in the league have been at least 7 feet tall. But even in this context, Wembanyama stands alone. Most 7-footers have been used as shot blockers; when they scored, it was almost always due to their extreme size. Wembanyama has mastered the skill sets of much smaller players. He can dribble through his legs; once I even saw him dribble through the legs of his defender. He can fluidly pull up and shoot from well behind the three-point line. It’s not a stretch to say that Wembanyama moves better with a basketball than anyone ever has at his height. “In all other instances, a 7-footer dribbling the ball up the court means that something has gone wrong,” the author and Spurs fan Shea Serrano told me. But when Wembanyama dribbles, Serrano finds it “good and right and holy.”

[Read: Air Jordan is finally deflating]

Wembanyama may possess preternatural grace for someone of his size, but he is still a coltish presence on the court. He seems to have stolen a taller man’s legs. Much of what he does comes easily on account of his enormous size, like a teenager having his way with a younger sibling’s Fisher-Price hoop. In a half-court offense, Wembanyama is never more than two (giant) strides from the basket, and when he arrives, he needs just a bunny hop to bring his forehead even with the rim. He can catch a lobbed ball with his back to the basket and execute a reverse dunk before he lands. For anyone else, these alley-oops would be spectacular, highlight-reel plays. For Wembanyama, they look like chin-ups.

As every sports marketer knows, identification is at the core of fandom. It is easier for us to bask in the glow of a great player if we can imagine ourselves executing their moves. Brands that endorse athletes count on people to buy into this fantasy of attainable greatness; they come right out and say so.

But it’s difficult for anyone to imagine doing the things that Wembanyama does, because he plays the game at such a high altitude. No normal person could ever match the skills of smaller NBA players either, but that fantasy is more accessible. Steph Curry, the league’s all-time record holder for three-pointers, has a degree of eye-hand coordination that is at least as freakish as Wembanyama’s height. But because Curry is just 6 foot 2, and I am nearly that, I can at least delude myself into thinking that with enough practice, I, too, could hit some of the shots that he does. It’s no accident that Curry, Michael Jordan, and other players whose physiques more closely resemble the everyman’s tend to have more fans, and more signature shoe lines. We look at Jordan and pretend that we can be like Mike.

Do people want to be like Wembanyama? Maybe so. He might be the front wave of a new era. The NBA may soon be stacked with even ganglier players who have all-world ball-handling skills and deep shooting range. Maybe my eyes will eventually adjust to them. Football fans needed time to accommodate themselves to Patrick Mahomes’s sidearm throws, as did the baseball fans who at first recoiled from Hideo Nomo’s tornado windup.

Either way, Wembanyama won’t be bothered. In his first season, he was named Rookie of the Year, and in his second, he would have been named best defender had he not been injured. This year, he looks even better, and Nike has already given him a signature shoe. A special logo is emblazoned on the heel and insole: an alien.

*Sources: Jim Poorten / NBAE / Getty; Chris Coduto / Getty; Stephen Gosling / NBAE / Getty; Adam Hagy / NBAE / Getty

The Black Hole That Could Rewrite Cosmology

2025-09-24T11:23:00-04:00

To study the origins of our universe is to struggle with profound chicken-or-egg questions. We know the Big Bang happened. Cosmologists can see its afterglow in the sky. But no one knows whether the laws of physics or even time itself existed before that moment. Nor can we say exactly what happened next. The order in which certain celestial objects formed during the very early universe is hotly contested.

For a long time after the Big Bang, not much of anything could form. All of space was permeated by a roiling plasma. It was too hot and chaotic for any structure to cohere. Hundreds of thousands of years passed before a tiny hydrogen atom could even hold itself together. Another 100 million years or so after that, great clouds of hydrogen condensed and stars flared into being. Most cosmologists believe that these stars were the first large, free-floating structures to illuminate our universe, and that black holes appeared later. But some have proposed that it went the other way around.

In the deep sky, beyond the most ancient fully formed galaxies, astronomers have now found a mysterious and colossal object that may be a primordial black hole. Earlier this month, a team of them posted an analysis of the object based on observations made by the James Webb Space Telescope. If their account holds up, the standard view of how the universe evolved will need serious revamping.

Long before black holes were ever glimpsed in reality, they were theoretical objects, products of the scientific imaginary. In 1783, the English natural philosopher John Michell proposed the existence of “dark stars,” objects of such concentrated mass that light cannot escape their gravity. Michell was reasoning from Newton’s laws. More than 100 years later, Karl Schwarzschild and Robert Oppenheimer brought his dark-star idea into alignment with Einstein’s theory of general relativity. They showed how an ultradense star could keep collapsing until space-time curved back on itself, sealing off its light in a black hole.

All of this work was done on chalkboards and in notebooks. Black holes would remain notional until 1972, when astronomers confirmed that they’d actually detected one. In the decades that followed, more of these exotic objects were found in every part of the sky. People have now seen small ones and big ones. They have picked up the tiny space-time ripples that emanate outward from two merging black holes. They have learned that most, if not all, galaxies have a black hole at their center. The supermassive one in the middle of the Milky Way shoots out jets of particles that expand into enormous bubbles. These bubbles appear to help regulate star formation and other cosmic processes here in the only galaxy known to host life.

[Read: When a telescope is a national-security risk]

Most of the black holes that astronomers have identified appear to be collapsed stars. But some theorists, including Stephen Hawking, have suggested that there might be other kinds in the universe. During inflation—an expansive process that theoretically took place just after the Big Bang—quantum fluctuations could have caused large parts of the cosmos to spontaneously buckle inward, forming black holes before any stars had yet appeared. But cosmologists have had trouble imagining the mechanisms that could generate such large fluctuations. If the mysterious object that the James Webb Space Telescope has found really is a primordial black hole, they will have to go back to their chalkboards and notebooks.

That we can even get a peek at something from the early universe is a technological miracle. The Webb telescope spotted this object way out in the dark realm beyond the last visible galaxies, where the only things that glow are likely proto-galaxies and other cosmic bits and bobs in various stages of formation. Even when black holes are close to us, they can be difficult to detect, because they trap light. To see a black hole, astronomers rely on the wrenching violence that it inflicts on nearby matter, which throws off sparks in the form of electromagnetic radiation. But if this object is a black hole too, then not much matter is surrounding it, so it isn’t throwing off so many sparks. (In cosmology terms, it’s nearly “naked.”) Most of what astronomers see in its vicinity is hydrogen and helium left over from the Big Bang—not what you’d expect from a black hole that had formed from a collapsed star.

[Read: Yes, a moon base]

We will need many more observations, and probably a larger space telescope, to know for sure whether it’s a primordial black hole. After all, our images of this object were taken from clear across the observable universe. They barely qualify as blurry snapshots, and the analysis of them hasn’t yet been peer-reviewed. Peter Coles, a theoretical cosmologist at Maynooth University in Ireland, has noted that the object might be some other kind of strange, celestial body instead. Other cosmologists suggested to me that it could be a black hole that formed directly from a gas cloud without having first become a star. It could be something else entirely. At the frontier edge of astronomy, tantalizing new observations have a tendency to be mirages.

We might learn that this one has been misinterpreted. We might find definitive proof that stars are older than black holes, just as cosmologists had long supposed. But even so, black holes would still retain some claim to ontological primacy, because they last so much longer. From their perspective, a star is just a transitory stage, a chrysalis. If the universe continues to expand as cosmologists predict, a day will come when star formation will cease altogether. Tens of trillions of years after that, the final stars will burn out. When that last stellar ember cools and darkens, the age of black holes will still be in its early days. Black holes will exist far, far longer than the entire illuminated age of stars. Of all the forms that this cosmos assumes, they will be among the most enduring. In a deep sense, this universe is theirs.

Are Humans Watching Animals Too Closely?

2025-09-07T07:00:00-04:00

Charles Darwin once noted that natural selection tends to preserve traits that conceal an animal in nature. It can paint camouflage onto their bodies with astonishing quickness: The peppered moth’s wings darkened only a few decades after England’s Industrial Revolution blackened urban tree trunks. Decades later, when pollution let up, their wings lightened again. But evolution has not moved quickly enough to conceal animals from human-surveillance technologies, which are undergoing their own Cambrian explosion. Cameras and microphones are shrinking. They’re spreading all over the globe. Even as we cause animals to dwindle in number, they are finding it harder and harder to hide.

Humans are closing in on a real-time god’s-eye view of this planet. Some subsurface places remain unmonitored. The sun’s light penetrates only a thousand yards down into the ocean. In the “midnight zone,” below that threshold, strange, glowing animals can still live a life of genuine mystery. But on the planet’s surface, humanity’s sensors are everywhere. Even animals in the Himalayas can be seen by the satellites that fly overhead, snapping color pictures. They can spot the hot breath of a single whale geysering out of its blowhole.

Deep in the wilderness, way off the hiking trails, scientists have laid out grids of camera traps. Automated environmental-DNA stations census animals in these places by gathering fragments of their genetic material straight from the air, or from veins in the watershed, be they trickles of snowmelt or full streams. The closer a landscape is to civilization, the more intrusively its animals are watched. Those that live in rural barns, feed lots, or aquaculture ponds are monitored by cameras. Along fence lines, their predators are too. Even herds that roam free on the open range are microchipped and trailed by drones.

Cities are the most potent nodes of this global animal panopticon. CCTV cameras stake out big public spaces, and Ring cameras peer out onto quieter streets. Smartphone-toting humans wander everywhere in between, taking geotagged photos of animals, including those in their home. They upload these images to social networks, hoping that they go viral.

Many animals appear to be entirely unbothered by all of this surveillance. Raccoons may show interest in a camera after it flashes, but then move on quickly. Birds have a mixed response: Black-tailed godwits seem to barely even register the nest cams that hover above their freshly hatched chicks. Other species are more likely to abandon a monitored nest. Some animals react even more strongly. The mighty tigers of the Nepalese jungle try to steer clear of camera traps, and at least one chimpanzee has executed a planned attack on a surveillance drone.

If animals do indeed have feelings about surveillance and privacy, those feelings won’t map cleanly onto ours. I recently had occasion to reflect on this while letting my dog, Forrest, out to relieve himself at night. I tend to watch where he goes in the yard so that when he’s done, I can call him right in and get back to bed. As a consequence, we sometimes make eye contact while he completes the act. It gives me an uneasy feeling, the green shine of his irises hitting mine just as his stream touches the grass. I wonder if my sleepy-eyed stare strikes him as intrusive.

I asked Alexandra Horowitz, who researches dog cognition at Barnard, if Forrest might be experiencing something akin to embarrassment during these moments. Horowitz, who has written multiple books about the mental lives of dogs, was reassuring on this point. (She would later have much more to say about the limited privacy that dogs are afforded.) She explained that dogs understand where people are looking, and that if mine wanted to hide his behavior, he would be unlikely to engage in eye contact. And anyway, in his olfactory social world, urination is a proud public act.

But all of this is speculation, Horowitz emphasized. We can’t ask animals directly whether they have their own notions of privacy, so we have to settle for these behavioral clues and the musings of philosophers. Since at least the 1960s, they have been asking whether animals might have privacy interests, and now that surveillance technology is spreading rapidly, a new generation has revived this question. Angie Pepper, a philosopher at the University of Roehampton, in the United Kingdom, answers in the affirmative. She points to animal behaviors that strongly suggest that some animals have privacy interests, including some that we are currently violating. She argues that coming to see these animals in a morally decent way may entail not seeing them at all.

There are some obvious ways that surveillance can harm animals. Animal-location data may be used for conservation purposes, but it can also be accessed by “cyberpoachers” or even the authorities. In 2014, an Australian government agency noticed that a GPS-tagged great white shark was swimming close to a beach and issued a kill order, even though the agency had no record of it ever approaching a swimmer. The order was withdrawn a week later, but had scientists never tagged the shark’s dorsal fin, it likely wouldn’t have been targeted by this precrime unit.

Just because surveillance might cause an animal harm doesn’t mean that its privacy has been invaded. But disturbing its tranquility might qualify, according to Martin Kaehrle, a Ph.D. student at the University of Wisconsin at Madison who has written about this subject. Many of our fellow creatures do seem to prefer feeling that some tiny corner of the universe is uniquely theirs, if only for a moment. When animals are packed together and deprived of that feeling, total social breakdown can occur. Pepper points out that pigs on factory farms commit acts of violence that would otherwise be rare in their communities. Some bite their neighbor’s tail without warning. Hens in similar situations will peck out one another’s eyes. In a famous experiment, a colony of mice was forced to live in tight conditions just so scientists could see what would happen. The colony quickly descended into indiscriminate violence, stopped mating, and died out.

Since at least the mid-2000s, birding groups have been passionately debating how best to preserve an animal’s tranquility, Kaehrle told me. He has spent years screenshotting these discussions on social-media sites, wildlife forums, and listservs. People argue about how much space a birder should give to its target, and whether baiting them with food is appropriate. Several communities agreed to implement total bans on location sharing.

In decades past, a birder who spotted a rare bird might notify someone at their local Audubon Society, who might then mark it with a colored pushpin on a map, or add it to a weekly recorded hotline message. Today, sightings flow much more quickly through digital-birding platforms, Discords, WhatsApp groups, and X accounts. One such account in New York City has tens of thousands of followers. A few years ago, the account doxxed a snowy owl, and it quickly became encircled by admirers, plus at least one drone. Snowy owls live in the High Arctic for half the year. By the time one reaches as far as New York, it is tired and hungry. If these endangered birds have to take flight over and over in order to avoid the boldest members of a human crowd, they can weaken further and even fail to mate.

Not all philosophers are willing to count these disturbances of an animal’s serene environment or personal space as an invasion of privacy. Some would argue that there are plenty of other reasons to think that harassing an animal is wrong. But a more straightforward case can be made in instances involving a more intimate kind of exposure. Humans are familiar with these scenarios, because we live in a complex social world, and we navigate it by presenting ourselves differently in different situations. You have a version of yourself who is the thinker of your innermost thoughts, the dancer before your bedroom mirror—but you likely present other versions in your interactions with your partner, kids, close family members, dear friends, doctors, and bosses. That’s why people don’t want their deepest secrets spilled onto the internet: Our ability to switch between selves would be seriously impaired. We would be forced into intimacy with everyone.

Many other animals also present different selves to different members of their communities. Kristin Andrews, a philosopher at York University, told me about gelada monkeys, which live in units consisting of one dominant male and about a dozen females. Gelada social norms dictate that the dominant male has sexual access to all of the females; a few follower males may be in the group but have no such access. When females mate with the dominant male, they do it out in the open and emit loud mating cries. It is a public act. But sometimes, for reasons that are her own, a female will transgress the community’s norms: She will seek to mate with another male, but not in public. The two will likely go for it when the dominant male is away, and they will emit much quieter mating cries.

Animal self-switching can also be detected in their communications. Some of their utterances are just indiscriminate broadcasts, but certain species use quiet tones to target a limited set of listeners, or even an individual. When humans communicate in this way, we reflexively describe it as private. Yet this has not stopped researchers from placing bioacoustic sensors in all kinds of wild habitats—and not only microphones: Seismic arrays of the sort that originally listened for nuclear tests have recently been used to detect the infrasonic rumbling of elephants. Teams of researchers are trying to use AI to decipher these rumbles.

Eavesdropping on elephants may not be technically possible, in the end. Either way, people probably won’t get too worked up about it, unless researchers use the information that they glean from an elephant wiretap to hurt the animals. But there is a class of animals whose privacy concerns are already acute: those that we keep in zoos or our homes. These animals are monitored by humans in ways that they likely would not choose. In zoos, many primates clearly prefer enclosures that give them the ability to retreat out of view. Not all of them get to make that choice. Neither do some of our most beloved pets.

“Dogs are given almost no privacy,” Horowitz told me. “I don’t know if they yearn for it, but in a typical home, they are expected to always be available. We even decide where they sleep.” Dogs don’t have a lot of opportunities for self-determination, Pepper told me when I asked her about pet privacy. “They always have to be accessible, not just in terms of sight but also touch.” Hearing this gave me a little jolt of shame. My Forrest is affectionate, but he is not a constant cuddler, like some of my previous dogs were. I probably force more hugs on him than I should.

We are not great respecters of boundaries, human beings. Dogs may not have known this about us when they first edged up to our campfires, more than 10,000 years ago. They could not have anticipated the degree to which we would dictate the most intimate parts of their lives, up to and including their sexual partners. Even after these dramatic interventions, which we have used to cultivate in dogs a preference for captivity, we still have to exercise a lot of coercion in order to get them to play along. We have to remove them from their mother while they are still young. We have to keep them behind locked doors and gates, and on leashes.

“It’s not obvious to me that the natural end point for dogs is this thick relationship where we dictate all aspects of their life,” Pepper said. “There are free-living dogs that have much thinner relationships with humans. They might stop by to get something to eat or to find somewhere to sleep, but they aren’t under this constant human control. Even the dogs that we have thoroughly socialized to live with us prefer varying levels of intimacy. Not all of them want to be with us all the time. They might seem like it when we come home at night, but in some cases, that’s because they didn’t have much company during the daytime.”

We can’t say what dogs’ preferences might be under different circumstances. But we do know that they have not chosen all of the intimacies that we impose upon them. They don’t get to decide the amount of distance that exists between them and us. They are expected to come right away when called. Rarely are they allowed to refuse our physical attention. There are moments when they may prefer to be untouched or unseen. Even when we are out of town, many of us watch them on cameras. We do all of these things because we love them, but this love is one that we thoroughly control. To them, at times, it may feel like something else.

Yes, a Moon Base

2025-08-10T07:00:00-04:00

No one can say that the Trump administration is entirely against alternative energy. In his first bold policy stroke as NASA’s interim head, Sean Duffy has directed the agency to put a 100-kilowatt nuclear reactor on the moon by decade’s end. This is not a lark. If humanity means to establish a permanent settlement on the moon, nuclear power will almost certainly be essential to its operation. And a lunar base may well be the most wondrous achievement in space exploration that people reading this will see during their lifetime.

The moon has gone unvisited, except by robots, for more than 50 years, and as of several months ago, it seemed as though Americans would be staying away from it for a good while longer. President Donald Trump was taking cues from Elon Musk, who seemed inclined to shelve the plan to put Americans back on the lunar surface and focus instead on an all-out sprint to Mars. But Musk has since fallen out of favor, and last month, congressional Republicans secured a funding boost for the moon program.

NASA astronauts are now scheduled to return to the moon in 2027, and if all goes well, they will be landing on it regularly, starting in the early 2030s. Each crew will carry parts of a small base that can grow piece by piece into a living space for a few people. The astronauts will also take a pair of vehicles for expeditions—a little rover that they can use for local jaunts in their space suits, and a larger, pressurized one that will allow them to go on 500-mile regolith road trips in street clothes.

A base on the moon would be more democratic than those that Musk and his acolytes have advocated building on Mars. Given shorter travel times, a greater number of people would be able to experience its otherworldly ashen plains. Their homesick calls to Earth would have only second-long delays, as opposed to minutes for a call from Mars.

[Read: Inside the Trump-Musk breakup]

But even a small encampment on the lunar surface is going to require considerable energy. Temperatures dip to –410 degrees Fahrenheit in the shade, and human bodies will need to keep cozy amid that deep chill. The International Space Station runs on solar power, but that won’t be enough on most of the moon, where nights last for 14 days. Some of the agency’s other off-world projects are powered by raw plutonium. Hunks of it sit inside the Mars rovers, for instance, radiating heat that the wheeled robots convert into electricity. These hot rocks are also encased inside NASA’s probes to the outer planets and their moons. Without plutonium, the two Voyager spacecrafts couldn’t continue to send data back to Earth as they recede from the solar system.

The moon base will need more than a radioactive rock. It will need a reactor that actually splits atoms, like the one that Duffy has proposed this week. Even if that reactor were to fail, the resulting meltdown wouldn’t present the same risks to humans that it would on Earth. The moon is already a radiation-rich environment, and it has no wind to blow the reactor’s most dangerous effluvia around; the material would simply fall to the ground.

Duffy framed his push to get the reactor in place as a matter of national security. NASA’s program to return to the moon, called Artemis, will be an international effort, with several countries contributing pieces of the final base. (Japan’s space agency has tapped Toyota to design the large, pressurized lunar vehicle.) But when the United States invited Russia to join, Vladimir Putin declined. He has instead opted to help out with a larger Chinese lunar base, which is supposed to include a nuclear reactor 10 times as powerful as the one that Duffy announced.

Last month, Bhavya Lal, who served as an associate administrator at NASA during the Biden administration and is now a professor at RAND, and her fellow aerospace expert Roger Myers released a report arguing that a country could sneakily establish a sovereign zone on the moon in defiance of the Outer Space Treaty just by building a reactor. For instance, the Chinese could insist on a buffer around theirs for the sake of nuclear safety, and use that to keep Americans away from desirable ice-rich craters nearby. Lal and Myers seem to have captured the new administration’s attention: Duffy’s new directive ordering the development of the reactor specifically mentioned this risk.

If worry over Chinese lunar land grabs is the motivation for a moon base, so much the better. Space exploration often requires a geopolitical spur. And if NASA can build this first small lunar settlement, something grander could follow close behind. Once the agency has mastered the construction of a 100-kilowatt lunar nuclear reactor, it should have little trouble scaling up to larger ones that can support tens, or even hundreds, of people—in bases of the size that now exist on Antarctica. Some space agencies have reportedly discussed building hydroponic greenhouses and other elaborate structures inside the voluminous caves that run beneath the moon’s Sea of Tranquility.

All of this infrastructure could enable some serious lunar dystopias. The moon’s surface could become an industrial hellscape, pocked with mining operations where robots and human serfs extract platinum and titanium for use in advanced electronics back on Earth. Or the Outer Space Treaty could break down and the moon could become a heavily militarized zone—even a staging ground for nuclear weapons.

But an inhabited moon could also be a global commons for research. Both the U.S. and China have developed designs for large radio telescopes on the lunar dark side, where they’d be shielded from Earth’s radio noise and would greatly aid the search for signals from distant civilizations. In one design, robots would spread a metal mesh from a crater’s center to its rim, turning its concave surface into a natural radio dish. One can imagine an astronomer at a lunar base, peering out from a porthole, seeing the Earth shining in the sky, picking out its individual oceans and continents, and knowing that on the moon’s opposite side, a giant ear would be listening for messages from other Earths and other moons, all across the Milky Way and far beyond.

Every Scientific Empire Comes to an End

2025-07-31T07:00:00-04:00

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

Roald Sagdeev has already watched one scientific empire rot from the inside. When Sagdeev began his career, in 1955, science in the Soviet Union was nearing its apex. At the Kurchatov Institute in Moscow, he studied the thermonuclear reactions that occur inside of stars. A few lab tables away, Andrei Sakharov was developing the hydrogen bomb. The Soviet space program would soon astonish the world by lofting the first satellite, and then the first human being, into orbit. Sagdeev can still remember the screaming crowds that greeted returning cosmonauts in Red Square. But even during those years of triumph, he could see corruption working its way through Soviet science like a slow-moving poison.

The danger had been present from the U.S.S.R.’s founding. The Bolsheviks who took power in 1917 wanted scientists sent to Arctic labor camps. (Vladimir Lenin intervened on their behalf.) When Joseph Stalin took power, he funded some research generously, but insisted that it conform to his ideology. Sagdeev said that his school books described Stalin as the father of all fields of knowledge, and credited the Soviets with every technological invention that had ever been invented. Later, at scientific conferences, Sagdeev heard physicists criticize the uncertainty principle of quantum mechanics on the grounds that it conflicted with Marxism.

By 1973, when Sagdeev was made director of the Soviet Space Research Institute, the nation’s top center for space science, the Soviets had ceded leadership in orbit to NASA. American astronauts had flown around the moon and left a thousand bootprints on its surface. Sagdeev’s institute was short on money. Many people who worked there had the right Communist Party connections, but no scientific training. Eventually, he himself had to join the party. “It was the only way to secure stable funding,” he told me when we spoke in June.

In 1985, Sagdeev briefly gained the ear of power. Mikhail Gorbachev had just become general secretary at 54, young for the Soviet gerontocracy. He promised broad reforms and appointed Sagdeev as an adviser. The two traveled to Geneva together for Gorbachev’s first arms talks with Ronald Reagan. But Sagdeev’s view of Gorbachev began to dim when the premier filled important scientific positions with men whom Sagdeev saw as cronies.

In 1988, Sagdeev wrote a letter to Gorbachev to warn him that the leaders of the Soviet supercomputer program had deceived him. They claimed to be keeping pace with the United States, but had in fact fallen far behind, and would soon be surpassed by the Chinese. Gorbachev never replied. Sagdeev got a hint as to how his letter had been received when his invitation to join a state visit to Poland was abruptly withdrawn. “I was excommunicated,” he told me.

Sagdeev took stock of his situation. The future of Soviet science was looking grim. Within a few years, government funding would crater further. Sagdeev’s most talented colleagues were starting to slip out of the country. One by one, he watched them start new lives elsewhere. Many of them went to the U.S. At the time, America was the most compelling destination for scientific talent in the world. It would remain so until earlier this year.

I thought of Sagdeev on a recent visit to MIT. A scientist there, much celebrated in her field, told me that since Donald Trump’s second inauguration she has watched in horror as his administration has performed a controlled demolition on American science. Like many other researchers in the U.S., she’s not sure that she wants to stick around to dodge falling debris, and so she is starting to think about taking her lab abroad. (She declined to be named in this story so that she could speak openly about her potential plans.)

The very best scientists are like elite basketball players: They come to America from all over the world so that they can spend their prime years working alongside top talent. “It’s very hard to find a leading scientist who has not done at least some research in the U.S. as an undergraduate or graduate student or postdoc or faculty,” Michael Gordin, a historian of science and the dean of Princeton University’s undergraduate academics, told me. That may no longer be the case a generation from now.

Foreign researchers have recently been made to feel unwelcome in the U.S. They have been surveilled and harassed. The Trump administration has made it more difficult for research institutions to enroll them. Top universities have been placed under federal investigation. Their accreditation and tax-exempt status have been threatened. The Trump administration has proposed severe budget cuts at the agencies that fund American science—the NSF, the NIH, and NASA, among others—and laid off staffers in large numbers. Existing research grants have been canceled or suspended en masse. Committees of expert scientists that once advised the government have been disbanded. In May, the president ordered that all federally funded research meet higher standards for rigor and reproducibility—or else be subject to correction by political appointees.

[Read: Trump’s ‘gold standard’ for science manufactures doubt]

Not since the Red Scare, when researchers at the University of California had to sign loyalty oaths, and those at the University of Washington and MIT were disciplined or fired for being suspected Communists, has American science been so beholden to political ideology. At least during the McCarthy era, scientists could console themselves that despite this interference, federal spending on science was surging. Today, it’s drying up.

Three-fourths of American scientists who responded to a recent poll by the journal Nature said they are considering leaving the country. They don’t lack for suitors. China is aggressively recruiting them, and the European Union has set aside a €500 million slush fund to do the same. National governments in Norway, Denmark, and France—nice places to live, all—have green-lighted spending sprees on disillusioned American scientists. The Max Planck Society, Germany’s elite research organization, recently launched a poaching campaign in the U.S., and last month, France’s Aix-Marseille University held a press conference announcing the arrival of eight American “science refugees.”

The MIT scientist who is thinking about leaving the U.S. told me that the Swiss scientific powerhouse ETH Zurich had already reached out about relocating her lab to its picturesque campus with a view of the Alps. A top Canadian university had also been in touch. These institutions are salivating over American talent, and so are others. Not since Sagdeev and other elite Soviet researchers were looking to get out of Moscow has there been a mass-recruiting opportunity like this.

Every scientific empire falls, but not at the same speed, or for the same reasons. In ancient Sumer, a proto-scientific civilization bloomed in the great cities of Ur and Uruk. Sumerians invented wheels that carried the king’s war chariots swiftly across the Mesopotamian plains. Their priest astronomers stood atop ziggurats watching the sky. But the Sumerians appear to have over-irrigated their farmland—a technical misstep, perhaps—and afterwards, their weakened cities were invaded, and the kingdom broke apart. They could no longer operate at the scientific vanguard.

Science in ancient Egypt and Greece followed a similar pattern: It thrived during good times and fell off in periods of plague, chaos, and impoverishment. But not every case of scientific decline has played out this way. Some civilizations have willfully squandered their scientific advantage.

Spanish science, for example, suffered grievously during the Inquisition. Scientists feared for their lives. They retreated from pursuits and associations that had a secular tinge and thought twice before corresponding with suspected heretics. The exchange of ideas slowed in Spain, and its research excellence declined relative to the rest of Europe. In the 17th century, the Spanish made almost no contribution to the ongoing Scientific Revolution.

The Soviets sabotaged their own success in biomedicine. In the 1920s, the U.S.S.R. had one of the most advanced genetics programs in the world, but that was before Stalin empowered Trofim Lysenko, a political appointee who didn’t believe in Mendelian inheritance. Lysenko would eventually purge thousands of apostate biologists from their jobs, and ban the study of genetics outright. Some of the scientists were tossed into the Gulag; others starved or faced firing squads. As a consequence of all this, the Soviets played no role in the discovery of DNA’s double-helix structure. When the ban on “anti-Marxist” genetics was finally lifted, Gordin told me, the U.S.S.R. was a generation behind in molecular biology and couldn’t catch up.

But it was Adolf Hitler who possessed the greatest talent for scientific self-harm. Germany had been a great scientific power going back to the late 19th century. Germans had pioneered the modern research university by requiring that professors not only transmit knowledge but advance it, too. During the early 20th century, German scientists racked up Nobel Prizes. Physicists from greater Europe and the U.S. converged on Berlin, Göttingen, and Munich to hear about the strange new quantum universe from Max Born, Werner Heisenberg, and Albert Einstein.

When the Nazis took over in 1933, Hitler purged Germany’s universities of Jewish professors and others who opposed his rule. Many scientists were murdered. Others fled the country. Quite a few settled in America. That’s how Einstein got to Princeton. After Hans Bethe was dismissed from his professorship in Tübingen, he landed at Cornell. Then he went to MIT to work on the radar technology that would reveal German U-boats during the Battle of the Atlantic. Some historians have argued that radar was more important to Allied victory than the Manhattan Project. But of course, that, too, was staffed with European scientific refugees, including Leo Szilard, a Jewish physicist who fled Berlin the year that Hitler took power; Edward Teller, who went on to build the first hydrogen bomb; and John von Neumann, who invented the architecture of the modern computer.

In a very short time, the center of gravity for science just up and moved across the Atlantic Ocean. After the war, it was American scientists who most regularly journeyed to Stockholm to receive medals. It was American scientists who built on von Neumann’s work to take an early lead in the Information Age that the U.S. has still not relinquished. And it was American scientists who developed the vaccines for polio and measles.

During the postwar period, Vannevar Bush, head of the U.S. Office of Scientific Research and Development under FDR, sought to make America’s advantage in the sciences permanent. Bush hadn’t liked the way that the U.S. had to scramble to staff up the radar and atomic-bomb projects. He wanted a robust supply of scientists on hand at American universities in case the Cold War turned hot. He argued for the creation of the National Science Foundation to fund basic research, and promised that its efforts would improve both the economy and national defense.

Funding for American science has fluctuated in the decades since. It spiked after Sputnik and dipped at the end of the Cold War. But until Trump took power for the second time and began his multipronged assault on America’s research institutions, broad support for science was a given under both Democratic and Republican administrations. Trump’s interference in the sciences is something new. It shares features with the science-damaging policies of Stalin and Hitler, says David Wootton, a historian of science at the University of York. But in the English-speaking world, it has no precedent, he told me: “This is an unparalleled destruction from within.”

I reached out to the office of Michael Kratsios, the president’s science and technology adviser, several times while reporting this story. I asked whether Kratsios, who holds the role that once belonged to Vannevar Bush, had any response to the claim that the Trump administration’s attack on science was unprecedented. I asked about the possibility that its policies will drive away American researchers, and will deter foreigners from working in American labs. I was hoping to find out how the man responsible for maintaining U.S. scientific dominance was engaging with this apparent slide into mediocrity. I did not receive a reply.

All is not yet lost for American science. Lawmakers have already made clear that they do not intend to approve Trump’s full requested cuts at the NIH, NSF, and NASA. Those agencies will still have access to tens of billions of dollars in federal funds next year—and blue-state attorneys general have won back some of this year’s canceled grants in court. Research institutions still have some fight left in them; some are suing the administration for executive overreach. Universities in red states are hoping that their governors will soon summon the courage to take a stand on their behalf. “Politically speaking, it’s one thing to shut down research at Harvard,” Steven Shapin, a science historian at the school, told me. “It’s another thing to shut down the University of Arkansas.”

The U.S. government doesn’t bankroll all of American scientific research. Philanthropists and private companies support some of it, and will continue to. The U.S. shouldn’t face the kind of rapid collapse that occurred in the Soviet Union, where no robust private sector existed to absorb scientists. But even corporations with large R&D budgets don’t typically fund open-ended inquiry into fundamental scientific questions. With the possible exception of Bell Labs in its heyday, they focus on projects that have immediate commercial promise. Their shareholders would riot if they dumped $10 billion into a space telescope or particle collider that takes decades to build and generates little revenue.

A privatized system of American science will be distorted toward short-term work, and people who want to run longer-term experiments with more expensive facilities will go elsewhere. “American science could lose a whole generation,” Shapin said. “Young people are already starting to get the message that science isn’t as valued as it once was.”

If the U.S. is no longer the world’s technoscientific superpower, it will almost certainly suffer for the change. America’s technology sector might lose its creativity. But science itself, in the global sense, will be fine. The deep human curiosities that drive it do not belong to any nation-state. An American abdication will only hurt America, Shapin said. Science might further decentralize into a multipolar order like the one that held during the 19th century, when the British, French, and Germans vied for technical supremacy.

[Read: ‘This is not how we do science, ever’]

Or maybe, by the midway point of the 21st century, China will be the world’s dominant scientific power, as it was, arguably, a millennium ago. The Chinese have recovered from Mao Zedong’s own squandering of expertise during the Cultural Revolution. They have rebuilt their research institutions, and Xi Jinping’s government keeps them well funded. China’s universities now rank among the world’s best, and their scientists routinely publish in Science, Nature, and other top journals. Elite researchers who were born in China and then spent years or even decades in U.S. labs have started to return. What the country can’t yet do well is recruit elite foreign scientists, who by dint of their vocation tend to value freedom of speech.

Whatever happens next, existing knowledge is unlikely to be lost, at least not en masse. Humans are better at preserving it now, even amid the rise and fall of civilizations. Things used to be more touch-and-go: The Greek model of the cosmos might have been forgotten, and the Copernican revolution greatly delayed, had Islamic scribes not secured it in Baghdad’s House of Wisdom. But books and journals are now stored in a network of libraries and data centers that stretches across all seven continents, and machine translation has made them understandable by any scientist, anywhere. Nature’s secrets will continue to be uncovered, even if Americans aren’t the ones who see them first.

In 1990, Roald Sagdeev moved to America. He found leaving the Soviet Union difficult. His two brothers lived not far from his house in Moscow, and when he said goodbye to them, he worried that it would be for the last time. Sagdeev thought about going to Europe, but the U.S. seemed more promising. He’d met many Americans on diplomatic visits there, including his future wife. He’d befriended others while helping to run the Soviet half of the Apollo-Soyuz missions. When Carl Sagan visited the Soviet Space Research Institute in Moscow, Sagdeev had shown him around, and the two remained close.

To avoid arousing the suspicions of the Soviet authorities, Sagdeev flew to Hungary first, and only once he was safely there did he book a ticket to the U.S. He accepted a professorship at the University of Maryland and settled in Washington, D.C. It took him years to ride out the culture shock. He still remembers being pulled over for a traffic infraction, and mistakenly presenting his Soviet ID card.

American science is what ultimately won Sagdeev over to his new home. He was awestruck by the ambition of the U.S. research agenda, and he liked that it was backed by real money. He appreciated that scientists could move freely between institutions, and didn’t have to grovel before party leaders to get funding. But when I last spoke with Sagdeev, on July 4, he was feeling melancholy about the state of American science. Once again, he is watching a great scientific power in decline. He has read about the proposed funding cuts in the newspaper. He has heard about a group of researchers who are planning to leave the country. Sagdeev is 92 years old, and has no plans to join them. But as an American, it pains him to see them go.

The Nuclear Club Might Soon Double

2025-07-08T05:45:00-04:00

Illustrations by Michael Haddad

Sign up for National Security, a newsletter featuring coverage of rising authoritarianism, military intelligence, and geopolitical conflicts.

Updated at 10:37 a.m. ET on July 12, 2025

Keiko Ogura was just 8 years old when the atoms in the Hiroshima bomb started splitting. When we met in January, some 300 feet from where the bomb struck, Ogura was 87. She stands about five feet tall in heels, and although she has slowed down some in her old age, she moves confidently, in tiny, shuffling steps. She twice waved away my offered arm as we walked the uneven surfaces of the Hiroshima Peace Memorial Park, first neutrally and then with some irritation.

Ogura can still remember that terrible morning in August, 80 years ago. Her older brother, who later died of cancer from radiation, was on a hilltop north of the city when the Enola Gay made its approach. He saw it shining small and silver in the clear blue sky.

Ogura was playing on a road near her house; her father had kept her home from school. “He had a sense of foreboding,” she told me. She remembers the intensity of the bomb’s white flash, the “demon light,” in the words of one survivor. The shock wave that followed had the force of a typhoon, Ogura said. It threw her to the ground and she lost consciousness—for how long, she still doesn’t know.

Like many people who felt the bomb’s power that day, Ogura assumed that it must have been dropped directly on top of her. In fact, she was a mile and a half away from the explosion’s center. Tens of thousands of people were closer. The great waves of heat and infrared light that roared outward killed hundreds of Ogura’s classmates immediately. More than 20,000 children were killed by the bomb.

[In Focus: Before and after the atomic bombing of Hiroshima]

Ogura told me that after the initial explosion, fires had raged through the city for many hours. Survivors compared the flame-filled streets to medieval Buddhist scroll paintings of hell. When Ogura awoke on the road, the smoke overhead was so thick that she thought night had fallen. She stumbled back to her house and found it half-destroyed, but still standing. People with skin peeling off their bodies were limping toward her from the city center. Ogura’s family well was still functional, and so she began handing out glasses of water. Two people died while drinking it, right in front of her. A black rain began to fall. Each of its droplets was shot through with radiation, having traveled down through the mushroom cloud’s remnants. It stained Ogura’s skin charcoal gray.

In the days following the bombing, Ogura’s father cremated hundreds of people at a nearby park. The city itself seemed to have disappeared, she said. In aerial shots, downtown Hiroshima’s grid was reduced to a pale outline. More than 60,000 structures had been destroyed. One of the few that remained upright was a domed building made of stone. It still stands today, not far from where Ogura and I met. The government has reinforced its skeletal structure, in a bid to preserve it forever. Circling the building, I could see in through the bomb-blasted walls, to piles of rubble inside.

Ogura and I walked to a monumental arch at the center of the Peace Memorial Park, where a stone chest holds a register of every person who is known to have been killed by the Hiroshima bomb. To date, it contains more than 340,000 names. Only a portion of them died in the blast’s immediate aftermath. Tens of thousands of others perished from radiation sickness in the following months, or from rare cancers years later. Every generation alive at the time was affected, even the newest: Babies who were still in their mothers’ wombs when the bomb hit developed microcephaly. For decades, whenever one of Ogura’s relatives took ill, she worried that a radiation-related disease had finally come for them, and often, one had.

[From the October 1946 issue: That day at Hiroshima]

As time passed, news that more countries had built nuclear arsenals reached Japan. Meanwhile, the hibakusha—the Japanese term for survivors of the nuclear attacks—were stigmatized as mutants. Ogura told me that girls in her summer camp looked for burn scars on her body in the shower. Some of her friends’ weddings were called off by prospective grooms who feared that birth defects would affect their future children. Ogura worried that her own wedding would be canceled right up until the ceremony.

Since the Hiroshima attack, Ogura and her fellow hibakusha have told and retold their stories of the bombing and its long aftermath. But even the youngest of them are now in their 80s, and soon they’ll all be gone. The horrific reality of an atomic attack is fading out of living memory—even as a new turn toward rapid nuclear armament makes the possibility of a full-blown nuclear war more likely.

For all the recent focus on Iran, in a cruel irony, East Asia is where the world’s fastest buildups are unfolding, in China and North Korea. A dangerous proliferation cascade may be about to break out, right in the shadow of Hiroshima. It would likely start in South Korea, and spread first to Japan. It might not stop there. The decades-long effort to keep nuclear weapons from spreading across the planet may be about to collapse.

One cold, windy morning in Seoul, a week before I met Ogura, I surrendered my phone at the gates of the Korea Institute for Defense Analyses, a government brain trust that advises South Korea’s Joint Chiefs of Staff. Inside the gray brutalist building, the nuclear strategist Heo Tae-keun was waiting for me. Heo had recently served as South Korea’s deputy defense minister for policy. In that role, he had led the country’s delegations in nuclear talks with the United States. He is a former brigadier general with a rugby player’s build, a sturdy presence in every sense. And yet, that morning, he seemed deeply troubled.

Michael Haddad

President Donald Trump had just begun his second term, but already he was showing less restraint than in his first. Almost immediately, he had threatened Denmark with military force, and he seemed content—delighted, even—to let Russia decide Ukraine’s fate. His disdain for old alliances unsettled Heo. “I am not sure what will happen in Trump 2.0,” he told me. In Korea, he said, in the cautious way of a diplomat, “he is perceived as more unstable in his decision making” than previous U.S. presidents.

[From the July/August 2022 issue: We have no nuclear strategy]

Stability is prized by nuclear strategists, who by dint of their profession have had to envision, with disturbing vividness, what instability looks like in the nuclear realm. As America’s dependability as an ally comes into question, Heo, like many other South Koreans, is looking around nervously at the dangerous neighborhood where his country is located. South Korea hangs like an earlobe off the eastern edge of Eurasia. Not even a tiny moat like the Taiwan Strait separates it from the three nuclear-armed autocracies immediately to its north. The first of them, North Korea, is still technically at war with South Korea, and Seoul’s 9 million residents are attuned to its closeness. From the city center, where skyscrapers stand alongside old palaces preserved since the Joseon dynasty, it takes just 40 minutes to reach the thin strip of land-mine-riddled wilderness that separates the two countries.

When North Koreans came pouring over the border at the start of the Korean War, in 1950, both peoples were poor, and still suffering the aftereffects of Japan’s brutal 35-year occupation. Then, for three years, that war raged up and down the peninsula, from snowy ridge to snowy ridge, killing more than 2 million people. Heo told me, laconically, that South Koreans have no desire to repeat that experience. He gestured toward the sleek, gleaming city outside his window. “We overcame the Korean War, and built an economy and way of life,” he said. North Korea has less to lose.

Kim Jong Un has ruled as dictator in Pyongyang for 13 years, during which he has often threatened the South with reunification by force, and, more recently, outright annexation, just as Vladimir Putin has attempted in Ukraine. Kim is quickly expanding his nuclear arsenal. He already has dozens of warheads, and has threatened to use them not only as defensive weapons of last resort, but in a first strike that would turn Seoul into a “sea of flames.”

[Uri Friedman: A third nuclear age is upon us]

For decades, the threat of intense U.S. retaliation helped keep Kim’s father and grandfather from invading the South. But Kim rules at a time when Pax Americana looks to be winding down. Under Trump, the United States is now reported to be considering pulling troops out of South Korea, though administration officials have denied that. “The Korean people do not know if the U.S. commitment to them is real,” Heo told me. They may soon decide that to deter Kim, they need nuclear weapons of their own.

For the better part of a century, the U.S. has sought to limit nuclear proliferation, with considerable success. American presidents have deployed diplomats, saboteurs, and brute military force to stamp out nascent nuclear-weapons programs in Latin America, Europe, Africa, and Asia. They have done so because nuclear weapons are dangerous, and because each new nuclear nation further dilutes the awesome power that America had when it was the only one.

Just once has the U.S. helped an ally start a nuclear-weapons program, by sharing technical research with the United Kingdom, its junior partner on the Manhattan Project. In 1946, Congress outlawed all such sharing, and in the decades since, U.S. presidents have worked to keep West Germany, Australia, Libya, Brazil, Sweden, and others from building arsenals—and even helped persuade South Africa to dismantle an arsenal that it had already built. Today, of the world’s 193 countries, only nine have nuclear weapons.

Left to its own devices, South Korea could easily have been the tenth. The country is wealthy and technologically adept, and with North Korea next door, it has sufficient motive. The reason the South Koreans don’t yet have an arsenal on hand is that both times they started to build one, an American president found out and persuaded them to stop.

The military junta that ruled South Korea in the 1970s launched the country’s first covert nuclear program after the U.S. signaled a pullback from Asia that would culminate in the fall of Saigon. The nervous generals were secretly negotiating with France to purchase a reprocessing plant. When Gerald Ford found out, his administration threatened to terminate the U.S.-Korean military alliance, and pushed to cancel the sale. In the end, South Korea ratified the Nuclear Non-Proliferation Treaty instead, in 1975.

Only six years later, after North Korea broke ground on a plutonium reactor, Ronald Reagan’s administration intervened to halt another such program. It was less serious than the first, but Reagan still wanted it canceled: He assured Chun Doo-hwan, South Korea’s president at the time, that U.S. ground troops would remain on the Korean peninsula indefinitely, and Chun agreed to shut weapons research down for good.

North Korea has not seen fit to restrain its nuclear ambitions in the same way. During the heady years after the Cold War, George H. W. Bush removed the American warheads that had long been stationed at bases in South Korea, then pressured its president to sign a joint pledge with North Korea to keep the peninsula forever free of nuclear weapons. That pledge proved to be a sham; North Korea tested its first crude nuclear device just 14 years later, during George W. Bush’s presidency.

Barack Obama, an optimist on all matters nuclear, believed that he could persuade China to lean on North Korea until it gave up its nuclear program. This didn’t work either. Chinese leader Xi Jinping’s first priority regarding North Korea was and is the stability of Kim’s regime: If Kim’s rule collapses, refugees will flood into China and Xi will lose the buffer state that separates it from South Korea, America’s longtime ally. Xi’s willingness to press was limited, and so Kim kept on building warheads.

Xi may feel, in any case, that he is in no position to lecture Kim about proliferation. He himself is engaged in the fastest warhead buildup undertaken by any country since the Cold War’s peak. For decades, China was fine with having a few hundred warheads on hand as a deterrent. But Xi is now adding about 100 a year. He wants an arsenal as large as the ones that the U.S. and Russia have, if not larger. It’s part of his Chinese Dream, the great rejuvenation that he has imagined for his country.

And so, in some sense, a destabilizing proliferation cascade has already begun in East Asia, and proliferation often begets proliferation. Julian Gewirtz, who served as the senior director for China and Taiwan affairs on the National Security Council during the Biden administration, told me that China’s astonishingly fast and ambitious nuclear buildup has unsettled countries all across Asia. In both South Korea and Japan, he said, these concerns, combined with uncertainties about the Trump administration, “may lead them to consider ideas that were once unthinkable.”

Kim is already estimated to have about 50 warheads, and the material needed to build as many as 90 more. His nuclear ambitions have grown along with China’s. He doesn’t want to be a nuclear peer of India and Pakistan, who have contented themselves with about 170 warheads each. Kim wants to have about 300, like the United Kingdom and France, sources told me.

Heo said that nuclear strategists have developed some notions about how Kim might use an arsenal of 300 warheads if nuclear war were to ever break out on the peninsula. The first 100 of them would likely be reserved for Kim’s short-range missiles. They would be able to reach targets in South Korea—military bases, airfields, ports, and perhaps even Seoul itself—in less than two minutes. The radius of the attack could then move beyond South Korea, with another 100 warheads available to strike the country’s regional allies, Japan in particular. Kim is trying to build reliable intercontinental ballistic missiles, onto which the remaining 100 warheads would be fastened. They could be launched all the way to the United States, in waves, to overwhelm missile defenses.

Michael Haddad

North Korea’s first ICBM test, in 2017, was a “threshold breaker,” Jake Sullivan, who served as national security adviser under Joe Biden, told me. It showed that Kim’s effort to build missiles that could reach the U.S. mainland was further along than previously thought. He may now be getting help from Russia, in exchange for the 14,000 troops and millions of rounds of ammunition that he has sent to Ukraine. If Kim could plausibly put Washington, D.C., or Los Angeles in existential jeopardy, would the U.S. really protect Busan and Seoul?

This is the question that haunts Heo. He knows that American presidents have a lot of wiggle room when it comes to protecting South Korea. The mutual-defense treaty between the two countries is vague. When President Dwight Eisenhower negotiated it, South Korea’s leaders were still eager to restart the Korean War, to defeat the North once and for all. Eisenhower was willing to station nuclear weapons in South Korea to reassure them, but he refused to promise American military support in every case of conflict between the two countries, because he feared that the South would deliberately provoke a war.

The U.S. has always been cagey about its nuclear contingency plans for the region. Even after North Korea acquired nuclear weapons, when Americans conducted tabletop exercises with South Korea, they would often end them just after North Korea launches its first missile, which is right when things get interesting, from the South Korean point of view. The United States Strategic Command, which operates America’s nuclear-weapons systems, doesn’t like to divulge its contingency plans. The South Koreans tend to “leak like a sieve, and their systems have been penetrated by the Chinese,” a former senior Pentagon official told me. STRATCOM officials have professed not to understand why South Korea should even require reassurance; their attitude was Our word has been good for decades, and it’s still good—just take it.

As Trump first rose to power, South Koreans found it more difficult to just take America at its word. In 2016, they watched in horror as he riled up rally crowds by denigrating America’s Asian allies as freeloaders. Trump said that South Korea and Japan were ripping off the U.S. in trade and sending only “peanuts” in exchange for an American military presence in the region. He seemed to take special pleasure in threatening to draw down, or perhaps even wholly remove, the nearly 30,000 troops stationed in South Korea.

During his first presidency, Trump flattered Kim, and flew to meet the North Korean dictator at summits in Hanoi and Singapore. In exchange for this sheen of legitimacy, Kim paused his missile tests, but only for a couple of years, during which he reportedly kept adding to his nuclear stockpile. A reminder of Trump’s failed policy can still be glimpsed from a border lookout point north of Seoul. When I visited it in January, I could see a pale-gray building a mile or so into the demilitarized zone, beyond wild bush and barbed wire. Trump and Kim met there in 2019, but since then, it has stood mostly vacant, a potent symbol of America’s newly unpredictable foreign policy.

According to opinion polls conducted in recent years, 70 percent of the South Korean public wants the country to have its own nuclear arsenal. In 2022, voters elected the conservative Yoon Suk-yeol, a hawk’s hawk on North Korea, to the presidency. Mira Rapp-Hooper, who served as the senior director for East Asia and Oceania on Biden’s National Security Council, told me that she and other officials grew concerned during Yoon’s campaign when he called for the return of tactical U.S. weapons to the Korean peninsula. After Yoon assumed power, the Biden administration tried to reassure him that no such arsenal was necessary. Biden’s staff proposed a grand gesture, a declaration that would serve as an addendum to the two countries’ vague mutual-defense treaty.

[Phillips Payson O’Brien: The growing incentive to go nuclear]

The Washington Declaration was announced during Yoon’s visit to the White House in April 2023. That night, at a state dinner held in Yoon’s honor, he and Biden clinked glasses to celebrate the 70th anniversary of the alliance. Yoon, who is not otherwise known for his personal charm, rose to the occasion, singing a few bars of “American Pie,” by Don McLean, in English, to loud cheers from the assembled guests. A few months later, an American Ohio-class nuclear submarine docked in Busan, as a show of strength. But by then, Biden’s presidency, and its policy of reassurance, was close to an end. Over the course of the following year, it became clearer that Trump would be his successor. For the second time in less than a decade, Americans would elect as their leader a chaotic and untrustworthy man who seemed hostile to the very concept of alliances.

When Heo and I discussed the possibility that South Korea may need to go nuclear, he emphasized that he wouldn’t want an arsenal just for its own sake. Members of the defense intelligentsia would prefer to keep the American alliance the way it is. But they have to prepare, in case South Korea is left to deal with Kim on its own. Like almost everyone I talked with in Seoul, Heo eventually mentioned Ukraine. When the Soviet Union fell, Ukraine had a nuclear arsenal on its soil, but Bill Clinton helped persuade the Ukrainians to give it up. Not to worry, he said. The U.S. will have your back.

Near the end of my time in Seoul, I sat down to lunch with Park Jin, who served as foreign minister under Yoon. We met at a café downtown, just as the morning’s snowfall was letting up. Park, 68, has the elegant manners that you might expect of a former top diplomat, and he was stylishly dressed in a black blazer and turtleneck, set off by a gray cashmere scarf. Just a few days earlier, in the hours following Trump’s inauguration, the new president had offhandedly referred to North Korea as a “nuclear power” in response to a reporter’s question about foreign threats. Park was focused on that remark. He told me he had initially hoped that it was a simple mistake, but those hopes were dashed when Trump’s incoming defense secretary, Pete Hegseth, used the same language in a written statement to the U.S. Senate.

[From the August 2025 issue: Tom Nichols on what it takes to launch a nuclear weapon]

This characterization may sound innocuous, given that everyone already knows that North Korea has a nuclear arsenal. But official recognition of a rogue nuclear power is usually a prize to be bargained for in geopolitics. It was not one that any previous American president had been willing to grant Kim, and certainly not for free. Park believes that Trump was using it as a concession to lure Kim to another meeting, one that could hasten his country’s abandonment by America. “The North Korea issue is the unfinished business from his first administration,” Park said. “And he’s a businessman.”

Having already conceded North Korea’s legitimacy as a nuclear power, Trump won’t have many cards to play if he does attempt another renegotiation with Kim. Now that Kim’s nuclear arsenal is larger and Russia is his ally, he has more leverage, and may not even wish to meet. In search of a deal, Trump might try to secure a commitment from Kim to stop building ICBMs that threaten the U.S., and then declare victory—leaving North Korea’s ability to nuke Seoul entirely intact. Several South Korean security elites told me that a deal like that would be tantamount to abandonment, especially if it were paired with a troop withdrawal.

During his first term, Trump asked his staff to set a troop withdrawal from South Korea in motion. James Mattis, his secretary of defense, reportedly slow-walked the request. Now, according to The Wall Street Journal, the Defense Department is reviewing its Korea policy, and a reduction in troops is being considered, although a Pentagon spokesperson denied that there was any “immediate” plan to draw down forces.

If Trump does try again to withdraw troops from South Korea, it’s not clear what would stop him. When Jimmy Carter attempted something similar, he was foiled by intelligence assessments that counseled strongly against it. But Tulsi Gabbard, Trump’s director of national intelligence, is an isolationist, and—like the rest of his Cabinet members—a loyalist above all else. She has already fired agents for an inconvenient intelligence assessment. She could make sure that no one stands in Trump’s way.

It can sometimes be helpful to think of there being two South Koreas. The country is highly, and maybe even dangerously, polarized. The month before I arrived in Seoul, Yoon had declared martial law on false pretenses. Shortly after I landed there, he was charged with insurrection. Walking the streets, I heard dueling mass protests, for and against him. A megaphone call-and-response boomed through the downtown high-rises. In early June, Lee Jae-myung, a liberal candidate, won the snap election to replace Yoon. Normally, the election of a liberal president would quell talk of a South Korean nuclear-weapons program for a while, but now even some of the country’s liberals are nuclear-curious. In March, two foreign-policy-establishment figures from the new president’s party said that it is time to consider nuclear armament.

Months before Trump’s reelection, Victor Cha, the Korea chair and president of geopolitics at the Center for Strategic and International Studies, sent a survey to hundreds of South Korea’s national-security elites. Of the 175 that responded, 34 percent said that they were in favor of South Korea acquiring its own nuclear weapons. But that poll is already out-of-date. The nuclear conversation among South Koreans has only grown louder since Inauguration Day, and Cha expects the volume to rise even more in the coming years. If a pro-nuclear consensus took hold among elites, it could all move quickly, because public support is already there, Cha told me.

I heard something similar when I visited Yang Uk, a nuclear strategist at the Asan Institute for Policy Studies, in Seoul. After giving me a tour of his office—a charmingly boyish space packed with model fighter jets and combat knives in glass cases—Yang told me that he, too, has been hearing more nuclear talk among South Korean strategists, and not only within the small clique that has long supported a homegrown nuclear program. It’s happening among lots of mainstream people, he said.

If South Korea were to launch a nuclear program, it would probably do so in secret. Its leaders would want to avoid suffering through an American-led sanctions regime, as India did after detonating nuclear devices in 1998. South Korea’s export economy would shrink rapidly if Hyundai and Samsung suddenly couldn’t sell their cars, smartphones, and chips abroad. “We would be fucked,” Yang told me. He may have been speaking personally: The Asan Institute is funded by an heir to the Hyundai fortune.

South Korea might secretly seek America’s blessing. Cha imagined South Korea putting a feeler out to the White House: You don’t have to support our nuclear program. Just don’t oppose it. Some people in the current Trump administration wouldn’t be inclined to oppose it at all. During his 2016 campaign, Trump himself suggested that South Korea and Japan should consider getting their own nuclear weapons. Elbridge Colby, now his undersecretary of defense for policy at the Pentagon, has said that the U.S. shouldn’t use sanctions to deter Seoul from developing them. Colby has just been put in charge of formulating America’s National Defense Strategy.

I called Scott Kemp to ask him how quickly South Korea could spin up a plutonium weapon. Kemp, a professor at MIT, is an expert on the industrial mechanics of proliferation who previously counseled the U.S. government on questions of this sort. He told me that in a mad-dash scenario, South Korea would probably need only a year to reprocess enough nuclear waste from its power plants to make a weapon. “There are plutonium-bomb designs floating around,” he said. “It would astonish me if South Korea had not acquired some of them.”

To build out an entire arsenal that would present a clear deterrent to North Korea would take longer, perhaps 10 years. “Those would be 10 very dangerous years,” Cha told me. Many of the riskiest scenarios introduced by nuclear weapons arise during these unstable “breakout periods,” especially when adversaries are operating with limited information.

If Kim learned of the program, he might use force to try to prevent its success, as Israel has in Iran. Even if he did not use nuclear weapons, he might try to invade, especially if there were fewer U.S. troops in his way. South Korea would be able to marshal a much more capable military response than Iran, and if a war did break out, it could last years and possibly draw in the neighborhood nuclear powers. Russia would probably back Kim, and China might pitch in too. In 2016, Xi Jinping levied harsh sanctions on South Korea just for installing a single missile-defense system. Xi would be aghast to learn that a new nuclear arsenal was materializing just 250 miles from the Chinese mainland. News of a South Korean arsenal would be consequential throughout East Asia. It would almost certainly spur further proliferation in North Korea and China, but also quite possibly in Japan.

Late one night after arriving in Tokyo, I met Cha for a drink on the top floor of the Okura Hotel. Beneath us, the city’s elevated freeways curved through a dense matrix of glass towers, giving the Akasaka district its layered and futuristic feel. Cha was in town for a security summit; in a ballroom on a lower floor, he and I had just attended a private speech by Shigeru Ishiba, Japan’s prime minister. Less than a minute into the speech, Ishiba had mentioned the threat from North Korea. Cha noted that for all of this public North Korea talk, in private, it was the prospect of South Korea going nuclear that seemed to spook Japanese security experts the most.

Japan and South Korea have mutual-defense commitments, but they are not friends. Koreans have not yet forgiven Japan for devoting an entire bureaucracy to the sexual enslavement of Korean women during its violent colonization of the peninsula. Japanese elites will tell you that their leaders have apologized many times for these crimes, and even paid compensation. Korean elites will tell you that the compensation was paltry, and the apologies heavy on the passive voice. They note that Japan’s history textbooks still take quite a sympathetic view of its imperial adventures in Korea.

Both countries depend on America for their national security, and neither wants to be the junior partner in the region. South Koreans do not like that the U.S. allowed Japan to reprocess uranium into plutonium, starting in 1987, while they still cannot. Japan’s conservatives wonder why it was South Korea that received a special Washington Declaration and not their country. You can imagine how tempers in Tokyo would flare if South Korea were to leapfrog plutonium-rich Japan and develop nuclear weapons first.

I asked Ken Jimbo, one of Japan’s most respected nuclear strategists, what his country would do in that instance. We met in a conference room at the International House of Japan, overlooking the institute’s famous garden. Originally owned by a samurai clan, it had, unlike most local Edo-style gardens, survived the Allied firebombing of the city. The red-and-white Tokyo Tower loomed behind it in the eastern sky. Jimbo told me that if South Korea built its own nuclear arsenal, the desire to possess such weapons would surely spill over to Japan. “We would have to be very serious about what to do next,” he said.

Japan has been rearming itself with impressive speed already. As the country’s war crimes have receded in historical memory and China has grown stronger, many Japanese have come to feel that the country’s pacifist constitution is outmoded. Jimbo told me that he was personally embarrassed when the troops that Japan sent to Afghanistan in 2001 weren’t allowed to join combat missions. During the decade following the outbreak of that war, Prime Minister Shinzo Abe led a movement to loosen the constitution. The country’s militarization has recently accelerated: By 2027, its defense budget will have surged by 60 percent in just five years. There isn’t yet a loud, open conversation about going nuclear in Japan, as there is in South Korea. As the lone people on Earth to have suffered nuclear attacks, the Japanese have so far remained committed to three “non-nuclear principles,” which require the country not to produce nuclear weapons, possess them, or host others’ on Japanese soil. A generation ago, belief in these principles was so strong in Japan that it was hard to imagine the country ever building an arsenal.

But antinuclear sentiment has lost potency during the past 20 years, according to Masashi Murano, a senior fellow at the Hudson Institute. China’s rapid nuclear buildup has unnerved the public, Murano said, and so has North Korea’s. Japanese media once covered Kim’s family as an eccentric sideshow. Now every smartphone in the country gets a push alert when Kim lobs a missile into the Sea of Japan, or over the Japanese archipelago and into the Pacific.

I asked Narushige Michishita, a strategist and professor at the National Graduate Institute for Policy Studies, in Tokyo, if he could imagine the circumstances that would push Japan to go nuclear. He told me that he would pay close attention to what America’s president did. I asked what kinds of things he would watch for. A map of East Asia sat unfurled between us. Michishita touched his finger to South Korea and Taiwan. If the U.S. abandoned either of them during a crisis, Japan would probably need to go nuclear, he said.

Scott Kemp, the MIT professor, told me that Japan has almost certainly already done the preparatory work. In 1969, Prime Minister Eisaku Sato is said to have signed a secret memo, instructing the government to make sure that Japan would be ready to build a nuclear arsenal should the need arise. That same year, Sato’s administration began to put an enormous amount of money into its nuclear program, which now reprocesses nuclear waste into plutonium. I asked Kemp how long Japan would need to make a single warhead. His answer: Only a month, if speed were of the essence.

Nuclear weapons can be thought of as a kind of cancer that started metastasizing through human civilization in 1945. A few times during the Cold War, this cancer threatened to kill off much of humanity, but a partial remission followed the fall of the Berlin Wall. The U.S. and the Soviet Union agreed on a limit of 6,000 deployed warheads each—still enough to destroy most of the world’s major cities many times over, but down from the tens of thousands that they’d previously stockpiled.

The high-water mark for the disarmament movement came in 2009, when President Obama called for a world without nuclear weapons. For this address, Obama chose Prague, the site of the Velvet Revolution. He cast his eyes over a crowd of thousands that morning, and then over the whole continent. Peace had come to Europe, he said. Now it was time to go further, and negotiate a new arms-control treaty with Russia. The very next year, the two countries committed to cap themselves at 1,550 deployed warheads. At the time, China still had fewer than 300. Disarmament wasn’t on the near horizon, but the trajectory was favorable.

How long ago that moment now seems. The world’s great-power rivalries have once again become fully inflamed. A year after invading Ukraine in 2022, Putin suspended his participation in the capping agreement with the United States. He has begun to make explicit nuclear threats, breaking a long-standing taboo. Meanwhile, the Chinese have slotted more than 100 ICBMs in deep desert silos near Mongolia. The military believes that the U.S. has to target these silos, and Russia’s silos, to deter both countries, and doing so eats up “a big chunk of our capped force,” the former senior official at the Pentagon told me. Nuclear strategists in both of America’s major parties are now pushing for a larger arsenal that could survive a simultaneous attack from Russia and China. Those two countries will likely respond by building still more weapons, and on the cycle goes.

The writer Kenzaburo Oe has argued that it is the Japanese—and not the American scientists at Los Alamos—who have most had to reckon with the possibility that all of these nuclear weapons could bring about our extinction, or something close to it. This national reckoning has a geography, and Hiroshima and Nagasaki are its sacred sites. The day before I met Keiko Ogura in the Peace Memorial Park, I rode the bullet train southwest from Tokyo past the snow-tipped cone of Mount Fuji, then Old Kyoto and Osaka’s outer sprawl. In the early afternoon, I arrived at Hiroshima station and made my way to prefecture headquarters to meet Hidehiko Yuzaki, governor of the Hiroshima prefecture.

Yuzaki’s warm cherrywood office is the size of a small apartment. He has been governor for more than 15 years, and in that time, he has become the global face of Hiroshima. He played a large part in the G7 meeting that the city hosted in 2023, and Obama’s official visit in 2016—the first by a sitting U.S. president. Yuzaki is sometimes criticized for what local rivals say is an excessive focus on international affairs, but he sees his work with foreign leaders as continuing a great tradition in Hiroshima, dating back to the second anniversary of the atomic attack on the city. The mayor at the time, Shinzo Hamai, organized a peace festival, and in a speech that afternoon, he argued that Hiroshima should take on a new role in global culture as a mecca for the contemplation of disarmament. Since then, the city has been rebuilt into a wholly modern metropolis, but also an open-air museum that forces the mind out of the abstract realm of grand strategy and into the concrete reality of nuclear war.

I asked Yuzaki if he has become disillusioned as the world has again tipped toward nuclear proliferation. Was he troubled that the fastest buildups are occurring in East Asia, in Hiroshima’s backyard? He told me that he was frustrated. It was disheartening to him that people hadn’t yet grasped the real meaning of nuclear weapons. So long as anyone has them, there is always a risk of proliferation cascades, and no one knows where this new local one may end. The desire for these weapons is contagious, and could spread well beyond nervous national-security types in Seoul and Tokyo.

Indeed, the entire Non-Proliferation Treaty regime could unravel altogether. When Israel, India, and Pakistan went nuclear, they were not part of the Non-Proliferation Treaty (nor are they today), but South Korea is a member in good standing and Japan is, in some sense, the treaty’s soul. If those two countries flout the agreement, it will have effectively dissolved. Jake Sullivan, the former U.S. national security adviser, told me that the risk of a global proliferation cascade would rise “considerably.” The initial regional cascades are easy to imagine. The American pullback in Ukraine has already made Poland and Germany a lot more interested in going nuclear. If the Iranian nuclear program survives Israel’s attacks and develops a weapon successfully, Saudi Arabia and Turkey will likely want arsenals as well. The number of countries that have nuclear arms could quickly double.

[From the November 2005 issue: William Langewiesche on how A. Q. Khan made Pakistan a nuclear power]

We have some muscle memory for how to manage nuclear rivalries among a few great powers, Sullivan told me. But a strategic landscape of 15 or 20 nuclear powers could be risky in ways that we cannot anticipate. The odds of a nuclear exchange occurring would rise. The most potent current warheads are more than 80 times as destructive as the bomb that leveled Hiroshima’s urban core, and they now fly on missiles that can reach their targets in mere minutes. It would take only one of them to all but erase Tokyo, Seoul, Beijing, or New York City. The total damage that even a limited exchange of these more powerful weapons would cause is mercifully unknown to us, but it may be vain to hope for a limited exchange. The most elaborate and significant war game in the literature suggests that the cycle of nuclear vengeance would continue until the arsenals of all involved parties are spent. If a nuclear conflict does someday break out, death and destruction might very quickly unfold on a planetary scale.

Every moment that humanity spends with these weapons spread across the Earth, pointed at one another, is a foolish gamble with the highest-possible stakes. We are betting every chip that our nuclear-weapons technology and alert systems will not malfunction in existentially dangerous ways, even though they already have, several times. We are betting that no head of state who has red-button access will descend into madness and start a nuclear war, even though we know that leaders run the whole gamut from Marcus Aurelius to Caligula.

Before I left the Hiroshima-prefecture headquarters, I asked Governor Yuzaki what people usually overlook when they come to his city. Yuzaki paused for a moment to consider the question. He has personally hosted heads of state who control these arsenals. He said that most people are moved. He has watched foreign dignitaries weep in Hiroshima’s museums. He has seen them sitting in stunned silence before the memorials in the Peace Park. People feel horrible about what happened here, he told me. But they don’t seem to understand that humanity is now risking something even more terrible. They think that Hiroshima is the past, Yuzaki said. It’s not. It’s the present.

This article appears in the August 2025 print edition with the headline “The New Arms Race.” It originally misstated which nuclear technology is used to reprocess nuclear waste into plutonium in Japan.

This Awful, Forgettable Heat

2025-06-24T15:10:00-04:00

Think of a famous storm—maybe Hurricane Katrina, gathering force over the warming Atlantic surface and pinwheeling toward the mouth of the Mississippi River to flood the great city of New Orleans. You may remember that Katrina killed more than 1,300 people. You may remember other, less deadly storms, such as Sandy, which killed dozens of people in New York City, and at least 147 overall. Now think of a famous heat wave. It’s more difficult to do. And yet, heat waves can be fatal too. In 2023, scorching weather lingered for more than a month in Phoenix, Arizona, pushing temperatures to 119 degrees and killing an estimated 400 people in the county. Two years later, it’s all but forgotten. A major storm is history. A major heat wave is the weather.

This week’s heat wave is menacing much of the entire country: Almost three-quarters of America’s population—245 million people—have been subjected to temperatures of at least 90 degrees, and more than 30 million people are experiencing triple digits, according to one estimate. Yet few of us will remember this shared misery, unless we ourselves happen to be hospitalized because of it, or lose someone to heat stroke. Instead, these few days will blur together with all the other stretches of “unseasonably warm weather” and “record-setting temperatures” that now define summer in America. They will constitute just one more undifferentiated and unremembered moment from our extended slide into planetary catastrophe.

Heat waves have always been anonymous disasters. They lack the flashy action of earthquakes, volcanoes, or plagues, and they don’t show up much in ancient histories and myths. No single heat wave from human history has been assigned the narrative resonance of the Vesuvius eruption, or the mythic power of the storms that imperiled Odysseus. When heat waves do appear in stories, they tend to come in aggregate, after a series of them, occurring over months or years, have intensified droughts and famines. Our main cultural record of these collected runs of extreme heat consists of ruins left behind by civilizations that vanished after too many rainless years and failed harvests.

What if heat waves could be called by name, like Katrina and Sandy? Maybe that would give them greater purchase on our cultural memory. Several organizations have recently argued that we ought to label heat waves as we do tropical storms. (This week’s, if it were the first in some new system, might be called “Heat Wave Aaron.”) Supposedly, this would make heat loom larger in public discourse: More people would become aware of it and stay indoors. In 2022, a team working with the mayor’s office in Seville, Spain, piloted this idea. They assigned a local heat wave that had reached 110 degrees the name Zoe. According to a paper the team published last year, the 6 percent of surveyed residents who could recall the name without prompting also said they’d engaged in more heat-safety behaviors.

No one knows whether that effect would have lasted through other heat waves, once the novelty of naming wore off for the Sevillians. Either way, the idea may be tricky to implement. In the Atlantic Ocean, fewer than 20 tropical storms, on average, are named each year. But the United States alone is subject to hundreds of annual heat waves, and they vary immensely in scale. Some are city-size, and others—like this week’s—drape themselves across the country like a thick and invisible down blanket. And unlike tropical storms, which are categorized according to wind speed, heat waves kick in at different temperatures in different places. (Seattle’s heat wave might be Santa Fe’s average summer day.) So which of these deserve a name tag, and which ones don’t? Even if the naming idea catches on, these details will need working out.

Alas, heat waves will likely remain anonymous for most of us for a good while longer, if not forever. But perhaps we should not be so ashamed of this. Our inability to record these sweltering spells in a more conspicuous way is shared by the natural world, which rarely shows the marks of an episode of hot weather in any lasting way. A storm or an earthquake can reconfigure a landscape in a single moment of violence, leaving behind scars that can still be seen with the naked eye millennia later. In nature, as in culture, heat waves tend to show themselves after they have piled up into a larger warming trend. Only then are they visible in tree rings and ice cores, in coastlines that move inland, and in the mass extinctions that glare out from the fossil record—a thought to console yourself with as you wait for this week’s heat to break.

The Nobel Prize Winner Who Thinks We Have the Universe All Wrong

2025-05-30T07:00:00-04:00

Updated at 11:02 a.m. ET on May 31, 2025

Adam Riess was 27 years old when he began the work that earned him the Nobel Prize in Physics, and just 41 when he received it. Earlier this year, Riess, who is now 55, pulled a graph-paper notebook off a bookshelf in his office at Johns Hopkins University so that I could see the yellowing page on which he’d made his famous calculations. He told me how these pen scratches led to a new theory of the universe. And then he told me why he now thinks that theory might be wrong.

For nearly a century, astronomers have known that the universe is expanding, because the galaxies that we can see around us through telescopes are all rushing away. Riess studied how they moved. He very carefully measured the distance of each one from Earth, and when all the data came together, in 1998, the results surprised him. They were “shocking even,” he told his colleagues in a flustered email that he sent on the eve of his honeymoon. A striking pattern had emerged: The galaxies were receding more quickly than expected. This “immediately suggested a profound conclusion,” he said in his Nobel Prize lecture. Something is causing the expansion of the universe to accelerate.

Riess’s genius lies in making precise observations, but the task of explaining the accelerating expansion that he discovered fell to theorists. They proposed the existence of dark energy: a faint, repulsive force that pervades all of empty space. The amount of dark energy that fits inside your bedroom, say, isn’t very strong. It won’t blow the walls out. But when dark energy’s power sums across truly cosmic volumes of space, it can drive galaxy clusters apart. And as this process puts more space between those galaxies, the repulsive force only strengthens, speeding up the expansion of the universe. Telescopes can see hundreds of billions of galaxies today, but trillions upon trillions of years from now, dark energy will have driven them all out of sight. Eventually, it will dilute every last bit of matter and energy into a cold equilibrium, a thin gruel of nothingness.

By doing the work that led to the discovery of dark energy, Riess had helped add the final piece to what has since come to be called the “standard model of cosmology.” Indeed, few people played a larger role in establishing the standard model as the field’s dominant theory of how the universe began, how it organized itself into galaxies, and how it will end. But in recent years, cosmologists, the people who study the universe on the largest scales of space and time, have begun to worry that this story, and particularly its final act, might be wrong. Some talk of revolution. A growing number now say that the standard model should be replaced.

Adam Riess is among them.

Whenever a big theory of the universe is teetering, the old guard tends to close ranks; hence, the classic joke about science progressing one funeral at a time. Riess easily could have joined the old guard. He could have been its commanding officer. When he returned from Stockholm with his prize in 2011, he found that his academic life had changed. People around him started to behave oddly, he told me. Some clammed up. Others argued with him about trivial things, he said, perhaps so they could boast of having dunked on a Nobel laureate. Riess was besieged with invitations to sit on panels, give talks, and judge science fairs. He was asked to comment on political issues that he knew nothing about. He told me he was even recruited to run major scientific institutions.

Riess wondered about that path—being the big boss of a NASA mission or gliding around a leafy university as its chancellor. He could see the appeal, but he hated fundraising, and unlike other, older Nobel laureates, he said, Riess still felt that he had scientific contributions to make, not as an administrator, but as a frontline investigator of capital-n Nature. “Scientists sometimes tell themselves this myth: I’ll go lead this thing, and then I’ll come back and do research,” he told me. But then, by the time they’ve finished up with their administrative roles, they’ve lost touch with the data. They become clumsy with the latest software languages. “The science passes them by,” Riess said.

Riess decided to stick with research. There was plenty to do. The standard model had not solved cosmology. Even in 2011, people knew that the theory was lacking some important details. For one, 96 percent of the standard model’s universe is made up of dark energy and dark matter—and yet no scientist had ever detected either one directly. Cosmologists had good reasons to believe that both exist in some form, but any intuitions about how one might find either in the actual universe had not proved out. Something major seemed to be missing from the picture.

To get a better handle on these mysteries, theorists needed some new data. They badly wanted to know the rate at which the universe expands at different times, and for that they had to know the distances to galaxies from Earth with greater precision. This was Riess’s specialty: He would wait until he saw a certain kind of star explode in a far-off galaxy, and then he’d photograph its unfolding detonation in real time. He knew these supernovas always reached a certain luminosity, which meant he could figure out how far away they were by measuring their brightness in his telescope. The dimmer they were, the farther away.

I’m making this sound a lot easier than it is. Taking a snapshot of an exploding star from tens of millions of light-years away involves many subtleties. You have to subtract out light from the bright stars that surround it, in its own galaxy. The glow of the Milky Way will also sneak into your images, and so will the sun’s; you have to get rid of those too. At the same time, interstellar dust clouds near the star will block some of its light, as will dust in the Milky Way. These dimming effects must be accounted for. The circuits and other parts of your telescope will add noise to your image. The hundreds of thousands of pixels in your camera aren’t all the same, and their differences will need to be sussed out ahead of every observation.

Riess had never stopped trying to master these delicate additions and subtractions of light. Within the field, his measurements have long been regarded as the most precise, according to Colin Hill, a cosmologist at Columbia who does not work with Riess. But in 2011, Riess and his team developed an even better technique for measuring cosmic distances with the Hubble Space Telescope. (The idea came to him in the swimming pool, he said.)

As these new and better data piled up, a problem soon emerged. With each measured distance to another galaxy, Riess would update his calculation of the current expansion rate of the universe. To his alarm, the answers he was getting differed from those produced another way. Some cosmologists don’t bother with the distances to galaxies and look, instead, at the afterglow of the Big Bang. They can then take the expansion rate that they see in that snapshot of the early universe and extrapolate it forward on the basis of assumptions from the standard model. In other words, the latter approach takes it as a given that the standard model is correct.

Riess expected that this discrepancy between the two expansion rates would fade with further observations. But it was stubborn. The more he looked at distant galaxies, the more pronounced the difference became. Indeed, the mere fact of its existence presented the cosmologists with a serious problem. They became so vexed that they had to give it a name: the Hubble tension.

Riess wondered if the observations of the early universe that fed into the other measurement’s equations might be wrong. But neither he nor anyone else could find fault with them. To Riess, this suggested that the Hubble tension could be a product of a broken theory. “It smelled like something might be wrong with the standard model,” he told me.

If the standard model were to topple, the field of cosmology would be upended, and so would an important part of the grand story that we’ve been telling ourselves about the end of the universe. And so, naturally, with weighty matters of career, ego, and the very nature of existence at stake, the Hubble tension has led to a bit of tension among cosmologists.

Some of the field’s most prominent scientists told me that they still expect the problem to disappear with more data, and that Riess may be getting ahead of himself. Wendy Freedman, a professor at the University of Chicago, has made her own measurements of the local universe, using different exploding stars, and the Hubble tension shows up in her data too. But it’s smaller. She told me it’s too soon to tell what the problem is: her measurements, the standard model, or something else. She would want to know the distances to many more galaxies before deciding on the culprit. She would also want to see multiple methods of measurement converging. At a minimum, hers and Riess’s should match up. Hill, the cosmologist from Columbia, expressed a similar view.

[Read: The most controversial Nobel Prize in recent memory]

David Spergel, the president of the Simons Foundation, who has for decades held a lot of sway in the field, agrees that it’s premature to start dancing on the standard model’s grave. “Adam speaks very loudly,” Spergel said. “He argues vociferously with whoever disagrees with him.”

Riess does indeed prosecute his case with vigor. Still, no one has been able to find an error in his measurements, and not for lack of trying. His numbers have been cross-checked with observations from both the Hubble and James Webb Space Telescopes. Sean Carroll, a cosmologist and philosopher at Johns Hopkins who is not on Riess’s team, told me that Riess has done a “heroic job” of knocking systematic errors out of his measurements. But Carroll said that it is still too early to tell if the Hubble tension will hold up, and definitely too early to throw out the standard model. “If the implications weren’t so huge, people wouldn’t be so skeptical,” Carroll said.

Riess grew visibly exasperated when we discussed these objections. He blamed them on the “sociology” of the field. He said that a clique of cosmologists—Spergel and “other graybeards”—who work on the early universe have tended to dismiss conflicting data. (For the record, Riess’s own goatee is observably gray.) Even so, at least one of them had come around to his view, he said. Riess had sent data to George Efstathiou, a well-respected early universe cosmologist who’d been a vocal skeptic of the Hubble tension. On his desktop computer, Riess showed me Efstathiou’s reply: “Very convincing!”

I didn’t want to make too much of what might have been politeness, so I followed up with Efstathiou myself. In the email that he wrote to me, he was more circumspect than he had been with Riess: “I don’t have much to say on the Hubble tension.” So far as he could tell, Riess’s measurements didn’t contain any errors, but he couldn’t rule out the possibility that something in them was wrong.

Riess believes that in time he will be vindicated. He believes that the Hubble tension will likely grow more pronounced and that more cosmologists will start to question the standard model. For someone who helped stand up that theory, he comes off as gleeful about this possibility. Maybe this is just his scientific mindset: always deferential to the data. Or perhaps he simply craves the thrill of being right, again, about the fundamental nature of the universe.

When I visited Riess, back in January, he mentioned he was looking forward to a data release from the Dark Energy Spectroscopic Instrument, a new observatory on Kitt Peak, in Arizona’s portion of the Sonoran Desert. DESI has 5,000 robotically controlled optic fibers. Every 20 minutes, each of them locks onto a different galaxy in the deep sky. This process is scheduled to continue for a total of five years, until millions of galaxies have been observed, enough to map cosmic expansion across time. The observatory was preparing to release its second batch of data. Riess thought the information might produce another challenge to the standard model.

In the simplest version of the theory, the strength of dark energy—the faint, repulsive force that’s everywhere in the universe, pushing it apart—is fixed for all eternity. But DESI’s first release, last year, gave some preliminary hints that dark energy was stronger in the early universe, and that its power then began to fade ever so slightly. On March 19, the team followed up with the larger set of data that Riess was awaiting. It was based on three years of observations, and the signal that it gave was stronger: Dark energy appeared to lose its kick several billion years ago.

This finding is not settled science, not even close. But if it holds up, a “wholesale revision” of the standard model would be required, Hill told me. “The textbooks that I use in my class would need to be rewritten.” And not only the textbooks—the idea that our universe will end in heat death has escaped the dull, technical world of academic textbooks. It has become one of our dominant secular eschatologies, and perhaps the best-known end-times story for the cosmos. And yet it could be badly wrong. If dark energy weakens all the way to zero, the universe may, at some point, stop expanding. It could come to rest in some static configuration of galaxies. Life, especially intelligent life, could go on for a much longer time than previously expected.

[Read: When a telescope is a national-security risk]

If dark energy continues to fade, as the DESI results suggest is happening, it may indeed go all the way to zero, and then turn negative. Instead of repelling galaxies, a negative dark energy would bring them together into a hot, dense singularity, much like the one that existed during the Big Bang. This could perhaps be part of some larger eternal cycle of creation and re-creation. Or maybe not. The point is that the deep future of the universe is wide open.

I called Riess after the DESI results came out, to see how he was feeling. He told me that he had an advance look at them. When he’d opened the data file in his office, a smile spread across his face. He’d been delighted to see another tough result for the standard model. He compared the theory to an egg that is breaking. “It’s not going to cleave neatly in one place,” he said. “You would expect to see multiple cracks opening up.”

Whether the cracks—if they really are cracks—will widen remains to be seen. Many new observations will come, not just from DESI, but also from the new Vera Rubin Observatory in the Atacama Desert, and other new telescopes in space. On data-release days for years to come, the standard model’s champions and detractors will be feverishly refreshing their inboxes. For the moment, though, Riess believes that the theorists have become complacent. When he reaches out to them for help in making sense of his empirical results, their responses disappoint him. “They’re like, Yeah, that’s a really hard problem,” he said. “Sometimes, I feel like I am providing clues and killing time while we wait for the next Einstein to come along.”

When I talked to Riess for the last time, he was at a cosmology conference in Switzerland. He sounded something close to giddy. “When there’s no big problems and everything’s just kind of fitting, it’s boring,” he said. Now among his colleagues, he could feel a new buzz. The daggers are out. A fight is brewing. “The field is hot again,” he told me. A new universe suddenly seems possible.

This article originally attributed to Riess the discovery that the farther away galaxies are, the faster they are receding. In fact, he found that galaxies are receding faster than expected.

Elon Musk’s Most Alarming Power Grab

2025-05-06T08:00:00-04:00

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

When Elon Musk’s engineers bundled a batch of prototype satellites into a rocket’s nose cone six years ago, there were fewer than 2,000 functional satellites in Earth’s orbit. Many more would soon be on the way: All through the pandemic, and the years that followed, Musk’s company, SpaceX, kept launching them. More than 7,000 of his satellites now surround Earth like a cloud of gnats. This fleet, which works to provide space-based internet service to the ground, dwarfs those of all other private companies and nation-states put together. And almost every week, Musk adds to it, flinging dozens more satellites into the sky.

I recently asked the space historian Jonathan McDowell, who keeps an online registry of Earth’s satellites, if any one person had ever achieved such dominance over the orbital realm, and so quickly. “This is unique,” he said. Then, after considering the question further, McDowell realized there was a precedent, but only one: Sergei Pavlovich Korolev, the Soviet engineer who developed Sputnik and its launch vehicle. “From 1958 to 1959, when no one else had any satellites in orbit, Korolev was the only guy in town.” Musk is not the only guy in town circa 2025, but the rapid growth of his space-based network may represent a Sputnik moment of its own.

Musk first announced his intention to build a space-based internet, which he would eventually call Starlink, in January 2015. He had plans to settle Mars, then the moons of Jupiter, and maybe asteroids too. All those space colonies would have to be connected via satellite-based communication; Starlink itself might one day be adapted for this use. Indeed, Starlink’s terms of service ask customers to affirm that they “recognize Mars as a free planet and that no Earth-based government has authority or sovereignty over Martian activities.”

Musk is clearly imagining a future in which neither his network nor his will can be restrained by the people of this world. But even now, here on Earth, space internet is a big business. Fiber networks cannot extend to every bit of dry land on the planet, and they certainly can’t reach airborne or seaborne vessels. More than 5 million people have already signed up for Starlink, and it is growing rapidly. (You may end up using Starlink when you fly United, for example.) In the not-too-distant future, an expanded version of this system—or one very much like it—could overtake broadband as the internet’s backbone. A decade or two from now, it could be among our most crucial information infrastructure. The majority of our communications, our entertainment, our global commerce, might be beamed back and forth between satellites and the Earth. If Musk continues to dominate the launches that take satellites to space, and the internet services that operate there, he could end up with more power over the human exchange of information than any previous person has ever enjoyed.

Musk recognized that Starlink’s early adopters would be in remote and rural areas, where cables may not reach, and there are few, if any, cell towers. The U.S. is, for now, his biggest market, and the U.S. government may soon become a major customer: President Donald Trump has just delayed a $42 billion federal effort to expand broadband services, especially in rural areas. His administration has decided to make that project “tech-neutral,” such that cable hookups aren’t necessarily preferred over satellite—which means that Starlink can compete for the money. In the meantime, Starlink’s internet service is now also in planes, in ships at sea, in deep jungles, tundras, and deserts. In Gaza, medics have used Starlink while healing the wounded. At times when the people of Myanmar and Sudan learned that the internet had been shut off by their autocratic governments, they turned to Starlink. Ukraine’s soldiers use it to communicate on the front lines.

Musk’s ability to deliver this crucial service—the ability to coordinate action in conflict zones—has given him unprecedented geopolitical leverage for a private citizen. Reportedly, Pentagon officials have already had to go hat in hand to Musk after he threatened to restrict Starlink’s service to Ukraine’s troops, who were using it to launch attacks inside Russia. “He is not merely a mogul,” Kimberly Siversen Burke, a director at Quilty Space, an aerospace-research firm, told me. “This is someone who can flip a switch and decide the outcome of a war.” (Neither Musk nor Starlink responded to requests for comment.)

[Read: When a telescope is a national-security risk]

Political leaders all over the world have come to understand that Starlink’s dominance will be hard to dislodge, because SpaceX is so good at making satellites and getting them to space. The company makes its satellites in a factory outside of Seattle. Even in their bundled-up, larval form, they are enormous. The newest ones weigh more than half a ton, and once their solar-panel wings unfurl, they measure about 100 feet across. The company can reportedly manufacture at least four of these behemoths a day, and SpaceX’s reusable Falcon 9 rocket can hold more than 25 of them at once, all folded up inside its nose cone. Musk is able to launch these bundles of satellites at a Gatling-gun pace, while his competitors operate at musket speed with rockets that must be rebuilt from scratch each time. Last year, SpaceX successfully lofted 133 rockets into orbit, and more than 60 percent of them were carrying Starlink satellites. Every one of Musk’s commercial competitors, and also every nation’s military combined, launched fewer rockets than he did.

Before the rise of SpaceX, the French company Arianespace had dominated the global satellite-launch market. But its newest rocket, the Ariane 6, has so far been a boondoggle, with development delays and a costly one-and-done design. (The company expects to launch only 10 of them a year.) This is one reason that Europe has had a hard time fielding a serious competitor to Starlink, despite a desire to reduce Musk’s influence on future conflicts on the continent. Europe is home to Starlink’s largest commercial competitor, at least to this point, in OneWeb, a subsidiary of the French company Eutelsat. OneWeb has more than 600 satellites, compared with Musk’s more than 7,000, and its hardware is less advanced. As a result, the internet service it provides is slower than Starlink’s.

Separately, European Union nations have spent years planning the construction of a dedicated network of satellites for military and civilian use. But this project was recently dealt a blow when Giorgia Meloni, Italy’s prime minister—a friend of Musk’s—announced that she now prefers a deal with Starlink. The governments of Germany and Norway are each working on their own sovereign fleets, but they’re nowhere near having them up and running.

[Read: The military is about to launch a constellation]

The U.S. government, too, would have good reasons to avoid full dependence on Musk’s company for access to the space-based internet. The American military has an orbital network of military-grade satellites that allows for secure government communications and reconnaissance. But this too is a Musk product: SpaceX builds the satellites and ferries them to orbit.

The Pentagon’s leaders know this is a problem, or at least they once did. During the end of the Biden administration, the U.S. Space Force published a new strategy that ordered policy makers to avoid overreliance on any single company. But that was before the Defense Department came under the control of Trump, whose victorious campaign received more than $250 million in support from Musk. When I wrote to the Pentagon to ask whether avoiding overreliance on one provider was still a priority, I did not hear back. Even if the agency does end up diversifying its vendors, that process will take years, Masao Dahlgren, a fellow at the Center for Strategic and International Studies who specializes in space and defense, told me. “You can look at the launch schedule, and look at how many you need up there, and tell that it’s going to be a while.”

China’s People’s Liberation Army reportedly has its own concerns about Musk’s dominance over the potential future of communication in space. Several Chinese companies are currently building satellite-internet services; the largest one has roughly 90 satellites in orbit at the moment, and provides service only in the city of Shanghai. If that pilot project works, the network’s operator intends to expand across the country and beyond. China’s total number of satellites could tick up fast, because unlike Europe, the country is actually capable of launching a lot of rockets.

But of all of the aspiring competitors to Starlink, the most formidable is based in the U.S. Although Amazon has only just started launching satellites for its Project Kuiper, the company is looking to manufacture several thousand more in the coming years. It has also done the hard work of designing small, inexpensive terminals for users on the ground, which can compete with Starlink’s sleek, iPad-size consumer equipment. If Jeff Bezos’s space company, Blue Origin, can make its own reusable rocket fully operational, Amazon will start flinging satellites up into the sky in big batches as SpaceX does.

Of course, Musk is not going to sit still while the rest of the space industry catches up. Starlink is already available in more than 100 countries, and in Nigeria, Africa’s most populous country, it will soon be the largest internet provider of any kind. Other developing countries will likely want to make that same leapfrog bet that they can skip an expensive broadband build-out and go straight to satellite. And not just for the internet: Musk recently secured permission from the FCC to offer cellphone service via Starlink too. And he’s doing all this with his current technology. If SpaceX can finish testing its much bigger, next-generation Starship rocket within a year or two, as analysts expect, Musk will be able to expand his orbital fleets dramatically. SpaceX has previously said that the Starship will be able to carry up to 100 satellites in a single launch.

“In five years, we’ve gone from around 1,000 functional satellites to around 10,000,” McDowell told me. “I would not be surprised if in another 10 years, we get to 100,000 satellites.” They will beam more information down to the Earth than those that whirl around it today. They will offer an unprecedented degree of connectivity to people and devices, no matter where they are on the planet’s surface. The space internet of the future may become the central way that we communicate with one another, as human beings. Information of every kind, including the most sensitive kinds, will flow through it. Whoever controls it will have a great deal of power over us all.

The ‘Profound’ Experience of Seeing a New Color

2025-04-23T15:30:57-04:00

The color “olo” can’t be found on a Pantone color chart. It can be experienced only in a cramped 9-by-13 room in Northern California. That small space, in a lab on the UC Berkeley campus, contains a large contraption of lenses and other hardware on a table. To see olo, you need to scootch up to the table, chomp down on a bite plate, and keep your head as steady as you can. A laser will be fired into one of your eyes, targeting more than a thousand of your cone cells. (The scientists will have mapped their location on your retina in advance.) The lasers will activate your color vision like nothing in the natural world: A small square of exotic color will appear, just off-center from the focal point of your vision, against a background field of gray. It may flicker a bit, depending on what’s happening with the contraption, but it will remain unmistakably there.

Austin Roorda, an optometry professor at Berkeley, may have been the first person in the world to experience the new color. At the very least, he was the second, he told me on a recent call. (Ren Ng, the other Berkeley professor who co-leads the research group with Roorda, told me that he thinks he went second.) Roorda’s name is listed 12th among the 13 authors of the academic paper, published last Friday, that announced the new color’s creation—or its discovery, depending on your philosophical view of the matter. I asked him how he’d qualified for such an honor. “I have a habit of wanting to be a subject in all experiments in my lab,” he said. It wasn’t an IMAX-scale spectacle, Roorda said. But even so, in the aftermath, he felt a kind of euphoria. He described the color as a beautiful, ultra-intense teal. “As a scientist, the experience was profound.”

James Fong, a Ph.D. student at Berkeley, was the first author on the paper—and the one to name olo—but he has never seen the new color. He and the other grad students on the project drew straws to decide who would get their retinas mapped, and he was unlucky. As a result, Fong has become afflicted by a very specific kind of color-blindness, which he wishes urgently to cure: When the next set of slots for retina mapping opens up, he will raise his hand, he told me. “I would be disappointed if I finished my Ph.D. program without seeing the color that I spent a good fraction of my time here studying.”

As a rule, scientists do not lead especially glamorous lives, especially not right now. But they do occasionally get to see extraordinary things before anyone else. Sometimes, this privilege is enjoyed—or seized—by one person, as when Howard Carter allowed himself a candlelit glimpse into King Tut’s unsealed tomb the afternoon before he fully surveyed its glittering wonders with Egyptian authorities. Other times, a whole group gets to revel in the exclusivity: The Soviet scientists who launched the Luna 3 probe in 1959 kept the first images of the moon’s dark side to themselves for days. The exclusivity period can run much longer than mere hours and days: More than 50 years passed from the moment when Jacques Piccard and Don Walsh saw the bottom of the Mariana Trench from the porthole window of their bathyscaphe to James Cameron’s return trip there in a high-tech submersible. These experiences are rarefied, but they do not always trigger joy. The astronaut Mike Massimino has said that he felt an extreme loneliness upon seeing Earth from space, because he could not share it with the people whom he loved most in the world.

Roorda had an inkling that his research team was going to see something special. For more than five years, they’d been trying to conjure up novel human experiences of color. They seemed to have a lead on others who work in the field. Human beings are visual creatures; sight is our primary sensory window onto the world. For most of us, the waking mental experience is dominated by a constant stream of color images generated by the eye and the cortex, but exactly how those images are constructed by the mind is not yet understood. By giving the visual system an entirely new stimulus—a color that does not exist in nature—Roorda’s team was hoping to tease out the different roles that the eye and the brain play in creating the cinema of our lives.

Verifying that Roorda and the other participants had indeed seen a novel color was tricky. Only one person witnesses the experience of color: the person who sees it. Philosophers have fretted about color’s inescapable subjectivity since the late 18th century, when John Dalton discovered red-green color-blindness—his own. (Dalton noted that a pink geranium looked dramatically different when viewed in broad daylight than in evening candlelight, and was astonished when his friends told him that they experienced no such effect.) Zed Adams, a philosophy professor at the New School who specializes in the experience of color, told me that many 20th-century philosophers were haunted by the idea that we’re all trapped in our own perceptual world. Everyone wants to believe that they see the true rainbow, but no one can be sure that they do, Adams said.

[From the February 2025 issue: The false promise of seasonal-color analysis]

The team at Berkeley eventually hit upon a way to confirm that the colors that were experienced by the five people who took part in the experiment were roughly similar. Their method was ingenious, according to Adams, who was not involved in the work. First, the team produced olo by targeting a specific set of color-sensing cone cells in each participant’s retina—the so-called M cones, which are never activated on their own in natural settings. Then, while the subjects were experiencing that color, they rated its intensity and performed image-matching exercises, comparing olo with its nearest natural cousins. This process couldn’t tell the researchers what olo was, in a deeper sense, but it did suggest that the five subjects of the study had experienced an image with roughly the same degree of color saturation. All five also described the color as being a sort of teal or a mix of blue and green. Whether they were experiencing the same teal or blue-green remains a mystery, of course.

Olo does not—and cannot—exist outside this peculiar technological setup, or one very much like it. In that sense, it’s a cyborg experience that a human being can have only with the help of a machine. But Fong hopes that it might be a first step toward enhancing human color vision in the everyday world. Scientists have already used gene therapies to add a third set of cone pigments to the retinas of male squirrel monkeys, which are born with only two. It appears to give the monkeys the ability to access new colors, although that research is not yet definitive.

Some human females have four types of cones in their retinas, instead of three. For most of them, the fourth cone doesn’t lead to richer color vision, at least not measurably. But there is one woman, perhaps the most famous research subject in all of color science—cDa29—who is able to distinguish among hues better than those of us who have three cones. Researchers in Roorda and Ng’s lab are trying to figure out whether this sensory superpower can someday be engineered into an adult human. They’ve been laying the groundwork by using their targeted lasers to mimic the patterns of retinal stimulation that a person with four kinds of cones would experience, but in people with three cones. If (and it’s a big if ) people’s brains are able to process the same fine color distinctions that cDa29 can, gene therapy could perhaps be used to add a fourth kind of cone to human retinas. In just weeks or months, they could find themselves in a new sensory world, populated by 10 times as many gradations of color.

[Read: What color is a tennis ball?]

Those kinds of treatments are still a ways off. In the meantime, Fong is having fun fielding all of the reactions to the olo paper. Reporters from all over the world have stormed into his inbox, demanding interviews. Many have asked to be rigged up to the machine in Berkeley to get a glimpse of the new color. Artists have also been in touch. (Maybe James Turrell could make use of this technology.) Stuart Semple, an artist from the U.K., has started taking preorders for a paint based on olo. It’s called “YOLO.”

Fong told me that he takes particular pride in having named the color. The name olo is a play on 0 1 0, which corresponds to the types of cone cells—the 1 is for M’s—that were stimulated to generate it. The team had considered all kinds of alternatives but agreed that his solution was the most elegant. Fong was delighted just thinking about it. “How many people have named a color?” he asked me. Fong said that he has come to love olo. He said he now prefers it to red, orange, yellow, green, blue, or purple. He might be the only person in the world who has never seen their favorite color.

About That ‘Possible Sign of Life’ on a Distant Planet

2025-04-18T10:59:00-04:00

Few forms of media can still grab the general public’s lapels and say, “The world has changed in an important way, and you should know about it, now” like a push notification from The New York Times. On Wednesday evening, a particularly enticing one from the Times flashed across millions of lock screens. “Astronomers detected a possible signature of life on a planet orbiting a star 120 light-years away,” it read. Soon after, The Washington Post followed up with a notification of its own, using similar language about a possible sign of life found on a distant planet called K2-18b.

The word possible is doing load-bearing—if not Atlas-like—work in these headlines: It is indeed “possible” that a team of astronomers led by Nikku Madhusudhan, a professor at the University of Cambridge, has found a “sign of life,” but only if a whole daisy chain of other possibilities turns out to be actualities. Despite these contingencies, Madhusudhan is quoted in the Times describing this moment—and, by implication, his own work—as “revolutionary.” But when I spoke with other scientists in the field, they were much more circumspect. Sara Seager, a professor at MIT who was once Madhusudhan’s graduate adviser, told me that when it comes to K2-18b, “enthusiasm is outpacing evidence.”

When you read that astronomers have detected possible signs of life on another planet, you might picture them all hunched over a blurry image of K2-18b, zooming in to the planet’s fine details. But no one has ever seen K2-18b in the round; we only know it’s there because its host star dims at regular intervals when the planet passes in front of it. Madhusudhan and his team used the James Webb Space Telescope to collect light from K2-18b’s host star during one of these transits by the planet. A tiny fraction of the light they collected shone directly through the planet’s atmosphere, and that light carried a record of the molecules that it encountered there all the way to the Webb. According to Madhusudhan and his team, one of the molecules that K2-18b’s atmosphere contained was dimethyl sulfide. On Earth, dimethyl sulfide is produced only by living organisms, and not just any organisms: It comes from algae, precisely the kind of life you might expect to find on a planet that is covered in warm oceans, as the team believes K2-18b to be.

It’s worth pausing to say that astronomers’ ability to tease these chemical records out of starlight is nothing short of miraculous. But it is tedious work, and not at all straightforward. (“Gas-phase photochemistry is complex shit,” one astronomer texted me.) The details are too technical to get into here, but the important thing to know is that Madhusudhan and his team did not directly detect dimethyl sulfide. The chemical is only one of several that could be responsible for the signal they found. And although it’s the most likely one according to their models, others disagree.

When I spoke with the astronomer Mercedes López-Morales, she noted that “revolutionary” discoveries in astronomy usually have a different publicity profile than this one. López-Morales is the associate director of the Space Telescope Science Institute, which operates the Webb Telescope on behalf of NASA, the European Space Agency, and the Canadian Space Agency. She pointed out that none of those agencies saw fit to send out so much as a press release about the finding, even though the work was carried out with their telescope. The only major press release came from the University of Cambridge, where Madhusudhan works. It touted the findings as the “strongest hints yet” of life outside the solar system. López-Morales thought this was overstated. If the result had been that solid, “I guarantee that we would have had the White House announcing this,” she said. (Madhusudhan did not respond to a request for comment.)

Even if the detection of dimethyl sulfide on K2-18b were to eventually prove out to the entire field’s satisfaction, its presence is not necessarily a bankable sign of life. Last year, the chemical was observed in the dead, icy spray of a comet. Whatever dimethyl sulfide’s organic origins might be on Earth, the universe, in all its creativity, is clearly capable of making it without the help of organisms. “If you really want to convince yourself that you’ve detected life, you probably want to see clusters of these biosignatures”—which is to say, other chemicals associated with life—David Kipping, an astronomer at Columbia University, told me. “And you want to have much more contextual information about the type of planet.”

We don’t have much contextual information about K2-18b at all. We don’t even know what kind of planet it is. When I asked López-Morales what we can say about K2-18b that is totally certain, she paused for a long while and then said, “We know there is a planet there; we know that.” Much else is a total mystery, or heavily contested. Madhusudhan pictures a “Hycean” world—wrapped in hydrogen, its surface hidden beneath a global ocean. Another group, studying the same data record, sketches a very different scene: a magma‑ocean hellscape, sterilized by heat. When I asked Michael Wong, an astrobiologist at the Carnegie Institution for Science, if this finding had, at the very least, made K2-18b the most interesting planet on which to look for life, he wouldn’t even go that far: “I’m not quite ready to say that.”

That scientists would want to generate excitement about their research among the general public is understandable. Many of them are funded by taxpayers, after all, and right now research budgets are being slashed, and future observatories are threatened with cancellation. But hype carries its own dangers. Several astronomers I spoke with for this story told me that friends and family had besieged them with text messages when the news broke. Judging by the volume of messages that I, too, received on Wednesday night, a great many people interpreted the Times’ push notification as announcing an epochal cosmic discovery. By the time I woke up yesterday morning, a backlash had already started brewing among scientists on Bluesky, and elsewhere. Ignas Snellen, an astronomer based in the Netherlands, told a reporter at the Dutch newspaper de Volkskrant that Madhusudhan’s framing of the research is “irresponsible nonsense.”

López-Morales worries that there could be a boy-who-cried-wolf effect. After seeing so many previous findings walked back, or dramatically qualified, people won’t believe a genuinely revolutionary discovery when it comes along. “We will one day find life on another planet,” López-Morales said. “That’s what so many of us are working towards. Statistically, with the number of galaxies and stars that are out there, I cannot imagine that there won’t be life elsewhere. But we’re not there yet.” If and when that push notification does light up our lock screens, it would be a shame if members of a jaded public decide to simply swipe it away.

Why Is Trump Mad at the Zoo?

2025-04-06T08:00:00-04:00

The order came down late in the evening, when the orangutans, lions, and crocodiles would be resting. The next morning, March 28, the animals awoke to a new political reality: The world’s most powerful man had taken an acute interest in their place of lodging, the National Zoo. President Donald Trump had directed Vice President J. D. Vance to rid the Smithsonian Institution of all “improper ideology.” As a ward of the Smithsonian, the zoo was not only covered by this mandate; it was specifically mentioned as one of the facilities to be cleansed of wrongthink.

Trump’s order leaves little mystery about what he wants changed at the Smithsonian’s National Museum of American History and its National Museum of African American History and Culture. It calls for the removal of “divisive,” “race-centered ideology” from those museums, and says that their exhibits should instead instill pride “in the hearts of all Americans.” But the order’s text is silent on the nature of the zoo’s ideological transgressions, and my email to the White House asking what they might be went unanswered. Trump has not previously been counted among the zoo’s critics, who tend to lament the life of captivity suffered by its animals, not their potential indoctrination.

I reached out to the zoo staff to ask if they knew what the administration wanted changed. When I did not immediately hear back, I decided to visit the zoo, in the mindset of a freshly appointed cultural commissar. One morning this week, I arrived at its Connecticut Avenue entrance. Pollen-coated cars were lined up outside, and blossoms on the zoo’s magnolias were turning themselves inside out in the clear morning sun. Just a few hundred yards down its central path, near the Asia Trail, a food truck was already serving cocktails.

[From the July 1919 issue: Pessimism and the zoo]

On my way over to the zoo, I’d read the institution’s most recent strategic plan. In the introduction, former National Zoo Director Steven Monfort says that by going from a global population of 1 billion to 8 billion in only 200 years, “humans have made things very hard for wildlife.” It occurred to me that Vance might find this characterization a touch too Malthusian; he has often railed against what he perceives as anti-natalism in liberal culture. But the sight of parents carrying Moscow mules and margaritas away from the food truck suggested family-friendliness, at least of a certain kind.

At the zoo’s newly renovated Bird House, I joined a long line of families clustered around strollers, waiting to be let into the aviaries. In 2023, I’d met the zoo’s chief curator for birds, Sara Hallager, while reporting a story about the institution’s decision to euthanize a fox that may have killed 25 of its flamingos. Hallager had told me that after the renovation, the zoo would no longer acquire birds from Africa, Asia, or South America. Its new exhibits would showcase only North American birds. Now I wondered: With this “America First” approach, had the zoo intended to obey (way) in advance? If so, that might explain why an enormous pink-marble sculpture of an eagle—salvaged from the original Penn Station—had been placed near the Bird House entrance.

As I moved deeper into the exhibit, this theory seemed less plausible. Its interpretive panels were not overtly political—I searched high and low for land acknowledgments and found none—but they also didn’t seem to have been designed to please Trump. For one thing, they’re printed in English and Spanish, a first for the zoo. They also celebrate the ability of migrating birds to move freely among the Amazon rainforest, North America, and the High Arctic.

I did find one potentially “divisive” panel in the turkey enclosure. It drew a distinction between North America’s Indigenous people, who hunted turkeys for thousands of years but took care not to wipe them out, and European colonists, who in just two centuries drove the birds to the brink of extinction. This may not be the sort of sentiment that “instills pride in the heart” of Americans. And yet it’s true.

[Read: The aftermath of a mass slaughter at the zoo]

Everywhere I went, I heard kids buzzing about the zoo’s new star attractions, two pandas named Bao Li and Qing Bao that Xi Jinping had sent from China as a gesture of friendship. A source at the Smithsonian Institution who was not authorized to speak to the press told me that before the pandas went on public view, the zoo had been besieged with messages from senators requesting advance meet and greets. I briefly entertained the thought that the zoo had ended up in Trump’s crosshairs because some key ally of his had been denied a picture with the bears. Whatever the case, Bao Li himself seemed entirely indifferent to politics. He sat, lolled back against a green hillside, chewing through whole sticks of bamboo like they were Twizzlers at the movies.

The zoo features less explicit climate advocacy than you might expect from an institution devoted to animal conservation. Most of it is concentrated in a single room in the Amazonia building. The Trump administration has been relentless about scrubbing government websites of all mentions of climate change, no matter how anodyne, but this was gentle stuff. In the center of a large mural from the 1990s recommending solar power, a kid wearing baggy clothes—now back in fashion—picks up trash in a forest. No fossil-fuel multinationals are named and shamed in the surrounding panels. The staff members in green vests did not appear to be indoctrinating anyone. They just gamely answered questions about the neon-blue tree frogs in a nearby terrarium.

The exit from Amazonia dumped me out onto a path that runs along the zoo’s southern edge. Traffic noise wafted down from the Duke Ellington Bridge, reminding me that I was not in a rainforest, but in the middle of Washington, D.C.—a city that Trump has derided as a “filthy and crime-ridden embarrassment to our nation.” Continuing down the path, I arrived at the Kids’ Farm exhibit, a shining scene of rural Americana that would not have been out of place on a butter label. Near the big red barn and stables, toddlers were perched on a fence, petting mules. A cow’s blotchy black coat shimmered in the bright heat of the afternoon. Like the Bootheel BBQ & Southern Catering food truck parked nearby, which promised to “feed your Southern soul,” the exhibit seemed designed to flatter, not antagonize, a narrow and nostalgic view of “real America.”

Before leaving the zoo, I popped into the visitor’s center. I confirmed that the bookstore inside was aimed at the nonpartisan animal lover, not the activist, and learned that the zoo usually holds a secular-coded celebration of Easter—its focus is nature’s post-winter bounty, not the newly risen Christ. The zoo’s website calendar does show that last year, and for several years prior, it also recognized International Family Equality Day. Local LGTBQ organizations participated in the event, and some described it as “Gay Day at the Zoo.” As part of the festivities, guests were able to watch a beaver or seal eat rainbow-dyed ice cake. Last year’s event also had a musical performance featuring themes of “climate justice, inclusion, queer identity, and community.” When I emailed the zoo to ask whether International Family Equality Day would continue this year, I did not receive a reply.

I could see how this celebration might inflame a social conservative, but the tame, one-day event did not seem like enough to merit the zoo’s inclusion in the executive order. Nor did any of the other things that I’d found—unless the administration is taking a “broken windows” approach to policing ideology.

Then again, I can’t claim that my audit was exhaustive. I had intended to visit every exhibit, but I ended up skipping the Reptile House. Not for lack of interest; it’s actually one of my favorite places at the zoo—the pythons and unblinking crocodiles provide a real encounter with the animal other. But the line was very long, with little shade. And so I can’t tell you for certain that the Reptile House isn’t a hotbed of critical race theory, or other MAGA heresies. Vance and his team will have to find out and let us know.

LeBron James and the Limits of Nepotism

2025-03-16T08:00:00-04:00

I need you to watch this 13-second video of ESPN commentator Stephen A. Smith walking to his courtside seat at a Los Angeles Lakers game. I need you to notice how Smith, perhaps the biggest voice in sports—in sheer decibels, if not reach—savors the see-and-be-seen pleasures of the courtside experience. That was two years ago. Now imagine how he might have floated into the Lakers’ home arena the night of March 6. Only hours earlier, it had been reported that ESPN had agreed to a new contract with Smith worth more than $100 million. To celebrate, Smith’s agent, Ari Emanuel, had invited him to the game. Larry David would join them courtside, near the end of the Lakers bench. This should have been the perfect atmosphere for Smith to revel in his ascent to the pinnacle of sports media. Instead, his night took a bad turn.

Early in the game, Smith made eye contact with Bronny James, a rookie reserve guard on the Lakers and, crucially, the son of the Lakers’ star LeBron James. Bronny has had a rough season. He has been dogged by accusations that he would not be playing in the NBA were it not for his last name. On the road, opposing crowds engage in mocking chants, begging the Lakers coach to put him in. In January, he had his worst game of the season; in 15 minutes, he missed all of his shots and turned the ball over three times. The next morning, on First Take, ESPN’s flagship morning show, Smith made Bronny the subject of one of his trademark rants. For more than four uninterrupted minutes, he pleaded with LeBron—as a father—to stop exposing his son. “Stop this,” he said. “Stop this.”

LeBron does not seem to have taken kindly to Smith’s unsolicited counsel. At the March 6 game, near the end of the third quarter, he approached Smith in his courtside seat. Looming over Smith in a manner that does not usually accompany friendly chatter, he barked something. His exact words can’t be heard in a fan video of the encounter, but his menacing tone is legible. According to Smith’s account of the exchange, LeBron said, “Stop fucking with my son. That’s my son.”

Smith did not respond to a request for comment. Neither did LeBron, so the details of what he said that night are unconfirmed. But even going by the body language and by related comments that LeBron later made on a hot mic, it does appear that his goal was to intimidate. By confronting Smith, LeBron sent a message not just to Smith but also to members of the wider NBA press corps, few of whom have Smith’s stature and influence: Criticize my son, and there will be consequences. Like many rich and powerful people, LeBron seems to want the fruits of global media stardom for himself and his kids, but not the corresponding scrutiny.

LeBron has himself to blame for the sad media spectacle of Bronny’s first season. In the months leading up to last year’s NBA draft, Bronny told reporters that it was his life’s great dream to be drafted by a pro team, any team. In off-the-record interviews, league executives made clear that because Bronny does not have his dad’s skill or size, he wasn’t all that draftable. In his short career at the University of Southern California, which was interrupted by a scary incident of cardiac arrest, he averaged only five points a game.

By all appearances, it was LeBron who made sure Bronny got drafted, first by puffing him up, saying publicly that his son was already better than many active NBA players. LeBron also made it known that he himself would play for whichever team drafted Bronny. (No prospect has ever been packaged so attractively, with the NBA’s first or second all-time best player as a throw-in.) On draft day, the Lakers selected Bronny in the second round. That night, Bronny shed tears; his dream had come true. At a press conference a few days later, the team’s coach, JJ Redick, tried to quiet any doubts. Bronny had earned the pick, he said, with hard work.

At the time, some of LeBron’s critics saw this as contemptible. I wasn’t one of them. Why appoint yourself to the meritocracy police just to make LeBron and his son your first arrests? LeBron’s own father was never in his life. He was still sleeping on friends’ couches late into his childhood. Parenthood means something deep to him; he has described playing with his son as the biggest accomplishment of a career that includes four NBA titles and four MVP awards. There was something beautiful about him using the leverage that he’s amassed in the basketball world to make that happen, by dragging his first-born boy into the league. Was it nepotism? Sure, but human life is shot through with all kinds of advantages of birth, and Bronny’s didn’t seem the most important or unfair.

LeBron deserves some empathy, even in the way he dealt with Smith. Nearly all parents experience something akin to his naive desire, a wish to give their kid the future of their dreams while shielding them from pain and disappointment. LeBron may have felt some anguish as he watched his son this season. Few things are as excruciating as watching your kids suffer. No amount of money or fame can insure against it. LeBron may well be haunted by his role in bringing about that suffering.

But even one of the greatest athletes of our age doesn’t—and certainly shouldn’t—have the power to protect his adult son from criticism. Indeed, what power LeBron has derives from the intense public curiosity about him as a basketball player and as a person. He was able to conjure a plus-one for a family member on an NBA roster precisely because he has been the league’s main character for a generation now. The media were always going to be interested in Bronny’s performance on the court, even if he’d made it there according to his own ability—and all the more so because he seemingly did not.

[Read: The secret code of pickup basketball]

LeBron has had more exposure to the NBA’s media ecosystem than any other active player. He knows that it rewards viral rants that lionize or denigrate a player based on their most recent game. And he should have known that Bronny would not be exempt from that dynamic, no matter how fiercely his father was protecting him.

Now, by appearing to threaten Smith, LeBron has not only acted like a petty strongman; he has drawn new attention to his son’s disappointing season, enlarging the very story that he sought to suppress. It’s a rare misstep for someone so media-savvy, who has amassed an enormous personal fortune while staying almost entirely scandal-free across a long career that began when he was still a teenager. The mistake is, perhaps, understandable. The emotions of parenthood are gigantic. They can knock anyone off their game, even the great LeBron James.

Illustration Sources: Tim Heitman / Getty; Gina Ferazzi / Los Angeles Times via Getty; Nathaniel S. Butler / Getty

Throw Elon Musk Out of the Royal Society

2025-03-12T07:58:13-04:00

With every passing day, it is harder to remember that Elon Musk was not always a political firebrand. The old Musk advocated for his business interests and professed to care deeply about climate change, but he largely stayed out of partisan politics. As a result, he was much more popular. He hosted Saturday Night Live and walked the Met Gala’s red carpet. He also received substantive honors, including election to one of the oldest and grandest institutions of science, the Royal Society. The fellowship put Musk in elevated company: In 2018, he traveled to London to add his signature to the society’s charter, alongside those of Albert Einstein, Charles Darwin, and Isaac Newton.

Thousands of scientists are now calling for Musk’s name to be blotted out from that charter’s fine vellum pages. The effort kicked off last summer, when 74 fellows (out of roughly 1,600) sent a letter to the Royal Society’s leadership, reportedly out of concern that Musk’s X posts were fomenting racial violence in the United Kingdom and could therefore bring the institution into disrepute. In November, one of the signatories, the neuropsychologist Dorothy Bishop, resigned from the Royal Society in protest of what she saw as inaction; her statement cited Musk’s derogatory posts about Anthony Fauci and the billionaire’s promotion of misinformation about vaccines. Then, last month, Stephen Curry, a biologist who is not himself a Royal Society fellow—or a guard on the Golden State Warriors—wrote an open letter calling for Musk’s expulsion, for many of these same reasons. It has since been signed by more than 3,400 scientists, including more than 60 actual fellows.

The society has not taken any disciplinary action in response to these entreaties. Musk himself made no comment on the campaign to oust him until March 2, when Geoffrey Hinton, “the Godfather of AI” (and a member of the Royal Society Class of 1998), lent his voice to the cause. In a reply to Hinton on X, Musk said that only “craven, insecure fools” care about awards and memberships. Despite this I’m-not-mad bravado, Musk seemed stung. He did not respond to emailed questions, but on X, he did accuse Hinton of cruelty. The following evening, the Royal Society convened a meeting to discuss the matter. It took place behind closed doors, and what transpired is still not entirely clear. (In an email to The Atlantic, the society said that all matters relating to individual fellows are dealt with in strict confidence.) According to one report, the society now plans to send a letter to Musk, though what it intends to write was undecided as of last week. At least for now, Musk’s fellowship seems to be safe.

This roiling at the Royal Society comes at a tricky time for scientific institutions, especially universities. Having perhaps waded too far into political disputes in recent years, the leaders of these institutions are now trying to stay out of politics at the precise moment when politicians are trying to damage them. Musk may have been spared, so far, by an understandable desire among the Royal Society’s leadership to stay neutral. Scientific organizations that succumbed to political orthodoxies, or enforced them, have often come to regret it. During the Cold War, some scientists in the United States faced professional penalties or outright ostracization because they were suspected of being Communists. In the Soviet Union, dissenting biologists were shipped to the gulag. At the peak of China’s Cultural Revolution, a physicist who worked on the “Western science” of general relativity could be charged with resistance and denounced at a public rally.

[Read: The death of government expertise]

These cases are extreme, but subtle interminglings of politics and science can be toxic in their own way. They can undermine the atmosphere of free inquiry that gives science its unique power, its ability to sift good ideas from bad in pursuit of a more expansive and refined vision of the universe. Even an institution’s well-meaning statement of support for a social cause may have a chilling effect on any member who does not happen to agree. A scientist’s success is determined in no small part by their peers’ appraisal of their work and character. Scientific institutions should therefore avoid actions that could be interpreted as political litmus tests. They largely do: No university would deny a Donald Trump–supporting grad student’s application for enrollment, at least not as a matter of official policy. And likewise, mere support for Trump should not and would not disqualify Musk from the Royal Society.

Of course, Musk’s support for Trump is not the issue here. In her resignation letter, Bishop raised the matter of his scientific heresies, specifically about vaccines, to argue that he breached the society’s code of conduct, which prohibits fellows from undermining the society’s mission. In 2021, Musk posted and later deleted a cartoon that depicted Bill Gates as a fearmongering villain who was trying to control people with COVID vaccines. In 2023, he insinuated that the NBA player Bronny James’s cardiac arrest could have been a side effect of those vaccines. Outrageous as these posts may be, Musk is allowed to be wrong about some things. Scientists are unevenly brilliant, if they are brilliant at all, and some of the best were heretics or even fools on one scientific issue or another. Lynn Margulis revolutionized evolutionary biology. She also promoted pseudoscientific theories of HIV transmission. Freeman Dyson had a better handle on the physical laws of the universe than almost anyone since Einstein, but he went to his grave a climate-change skeptic. Kicking Margulis and Dyson out of polite scientific society for these consensus violations would have impoverished science.

[Read: The erasing of American science]

The best case for booting Musk from the Royal Society doesn’t concern his beliefs at all. It proceeds from his actions, the way that he is degrading the world of science on Trump’s behalf. In the months since the 2024 election, he has made himself into a tool of Trump’s administration, a chain saw, in his own telling. And with that chain saw, the president has begun dismembering America’s great scientific institutions. The Royal Society is an ancestor of those institutions. During its centuries-long heyday, it funded scientific research that wouldn’t otherwise have been pursued. The National Science Foundation and the National Institutes of Health do this today, on a much larger scale. The Royal Society’s members are right to feel a twinge of solidarity as they watch Musk and his Department of Government Efficiency push deep staff cuts at the NSF and the NIH, and hear reports that deeper cuts are to come. In their speed and extent, these force reductions have no precedent in the history of American science.

Musk has done all of this in the name of efficiency, but scientific research is antithetical to unrelenting thrift. Basic research needs some slack to allow for false starts and trial and error. Musk of all people should know this. When the Royal Society announced that Musk would be made a fellow, it cited SpaceX’s advances in rocketry, first and foremost, and rightfully so. The company has made reusable rockets a reality, and if its larger Starship model starts working reliably, it will enable a host of new wonders in space. But SpaceX’s success required long experimental phases and lots of exploded rockets, all of which cost money. A big chunk of that money came from NASA, a scientific institution whose checks are signed by the U.S. taxpayer. Last week, it was reported that NASA, too, will soon face budget cuts. They are said to be concentrated in its science division.

Maybe Musk values scientific institutions only as a means to his personal ends. Maybe he sees them as disposable ladders that he can cast off after he has climbed to new heights of wealth. Either way, scientists are finding it difficult to fight back. They don’t have much money. Their petitions and open letters can be cringeworthy. But at the very least, they ought to withdraw the honors that they have extended to Musk. They don’t have to let him retain the imprimatur and gravitas of the Royal Society, one of the most storied institutions to have come down to us from the Enlightenment. If you give a man a medal, and he returns with a torch to burn your house down, figure out how to stop him, fast. But also: Rip the medal from his chest.

The Golden Age of Antarctic Science May Be Ending

2025-02-28T14:05:34-05:00

Updated at 3:15 p.m. ET on February 28, 2025

Field research in Antarctica is an extended exercise in endurance. Grant approval alone can take more than a year, long enough to grow a beard like one of Ernest Shackleton’s shipwreck survivors. The military cargo planes that carry researchers from New Zealand to McMurdo, America’s largest coastal base on the continent, regularly turn back halfway in bad weather. At McMurdo, scientists sometimes wait for weeks for a safe flight into the frozen interior. They land on sketchy, unimproved runways and make camp right on the ice sheet. Scientific instruments occasionally glitch and break down in the extreme cold, and so do human bodies. Just when everything finally starts working, a rogue storm may roll over the horizon. If things go sideways, a rescue flight might be a week away.

President Donald Trump has just made this work even more difficult. Last week, his administration initiated a round of coarse-grained layoffs at the National Science Foundation (NSF), the federal agency that funds and supports all American field research in Antarctica. The team of program directors that shepherds this research to fruition was already shorthanded, and now it’s down to just one permanent full-time staffer. The approval of new research proposals will slow considerably, and in the coming years, the amount of research conducted each field season could be severely reduced. As far as anyone knows, these firings were not deliberately designed to derail American research in Antarctica. But they are nonetheless part of a larger dismantling of climate science across the federal government. Whether through sheer carelessness or in pursuit of his stated belief that climate change is a “hoax,” Trump is sabotaging this science when it is more crucial than ever to the human future.

The U.S. has long had the most extensive system of bases and aircraft in Antarctica. The NSF has an unmatched record of safely placing researchers in the continent’s most forbidding environments, and not just for intellectual pursuits. They are on urgent business, trying to find out just how quickly Antarctica’s ice sheets will crumble and melt into the ocean, raising sea levels high enough to drown coastal cities worldwide. These scientists drill deep holes through the mile-thick ice to see what’s happening at a glacier’s base. They send submersibles into the crucial zone underneath ice shelves, where warm water eats away at the frozen support structures that stop glaciers from sliding faster into the sea. They measure the annual ice loss on Antarctica, which directly correlates with sea-level rise worldwide. They put all these data in context by pulling up cores of ice that give us a million-year look back at the continent’s climate.

Despite the civilizational import of this work, in recent years, Antarctic research has struggled. At full strength, the NSF’s Antarctic Sciences Section is supposed to have approximately nine full-time program directors keeping the field research machine running. These directors evaluate research proposals. They assemble expert panels to review them. They go to Antarctica to support the fieldwork that they approve. Even before Trump took office, funding constraints and a long pandemic hangover had slowed polar research. Fieldwork was regularly being reduced in scope or postponed, sometimes for years. The fleet of C-130s that carry scientists to the poles had started to age badly. The agency demolished a dorm at McMurdo years ago, and still hasn’t rebuilt it. In December 2024, only four full-time permanent program directors, plus one part-timer, were in place for Antarctic science.

Of those four, three were in their probationary period, for one reason or another, making them vulnerable to the Trump administration’s order laying off probationary federal employees. These firings weren’t carried out with a great deal of grace. At 9:30 a.m. last Tuesday, David Porter, a program director, received an ominous email requesting his presence at a Zoom meeting. The meeting was a mess, Porter said. The agency’s staff had sent out a faulty Zoom link, and at one point, Porter was accidentally made a host. The director of the NSF, Sethuraman Panchanathan, was conspicuously absent. (The NSF did not respond to a request for comment on Panchanathan’s absence. In an emailed statement, an NSF spokesperson acknowledged that the organization had terminated 170 employees and said, “We thank these employees for their service to NSF and their contributions to advance the agency mission.”) Kelly Brunt, another program director, received word about the meeting while she was still on her way home from Antarctica after dropping off her big red NSF-issue parka in New Zealand. Her flight to the U.S. made an unplanned landing on the way, and she had to log on to the meeting from her hotel room before dawn, bleary-eyed and anxious. “Right away, we were told that we would be terminated by the end of the day,” she told me.

Now only one full-time permanent program director remains, along with a few part-timers and staff that rotate in from academia. There is no indication that the agency will be able to hire more. The remaining members of this patchwork crew will have to keep working through their own backlogs. They’ll attend to the laid-off directors’ proposals and half-finished projects if and when they can. Porter told me that in the final hours during which he could access his email, he contacted researchers whose projects he was supporting. “All I could tell them was, ‘Hey, you aren’t going to hear from me again, but good luck.’”

[Read: The erasing of American science]

The polar-research community is pessimistic about the odds of recovery. Ted Scambos, a glaciologist at the University of Colorado at Boulder, fears that funding cuts will follow the loss of the program directors. “There are a few senators who are champions of this stuff, but in this budget-cutting environment, they are going to have lots of other priorities to protect,” he told me. Sridhar Anandakrishnan, a glaciologist at Penn State, told me he worries that the global prestige of America’s polar-science programs will suffer. Work of this kind requires permanent infrastructure that only rich countries can build. The NSF has recently sought to make up for its various problems by leaning into international partnerships, but other countries may not be as interested in them going forward, Anandakrishnan said. “They may wonder if the NSF even has the bandwidth, or if it can still be trusted as a partner.”

Trump has always recoiled from the international character of climate science, and since his return to the White House, he has sought to hobble its global institutions. On his first day in office, he withdrew the U.S. from the Paris Agreement. He later forbade American government scientists from working on the Intergovernmental Panel on Climate Change (IPCC). On the domestic front, Trump has frozen all NSF grants, including those that fund climate science. A court has since paused that order’s enforcement, but some money transfers are reportedly still being held up. A wait-and-see atmosphere has seized the whole field while Congress negotiates a new federal budget: Some graduate programs have paused admissions and universities have instituted hiring freezes. The Trump administration has reportedly proposed a 30 percent budget cut to the National Oceanic and Atmospheric Administration, one of the world’s most comprehensive sources of climate data. Mass layoffs at NOAA have reportedly already begun. The administration did many of these things during the planet’s hottest January on record.

Climate change is a global problem, and its effects are best studied closely wherever they occur, but especially in Antarctica. More than 80 percent of the planet’s ice sits atop its surface. Scientists have lately focused on West Antarctica’s ice sheet, because it has historically looked more unstable. If and when it slides into the ocean, sea levels could rise by up to 10 feet, erasing islands and inundating cities that sit near the world’s shorelines, where more than 40 percent of all people live. Anandakrishnan told me that he and his colleagues now suspect that East Antarctica’s ice sheets could be similarly unstable. It will take work to find out how shaky they are, and whether anything can be done to keep them in place.

[From the July/August 2024 issue: A wild plan to avert catastrophic sea-level rise]

Trump will almost certainly bequeath a warmer Earth to the next administration. The past five American presidents have each done the same. What makes Trump different is his insistence on disrupting the basic apparatus of climate science, the effort to understand how warm the planet will one day become, and how quickly its seas will rise. Heating up the world for future generations is a crime, but this is a cover-up.

If DOGE Goes Nuclear

2025-02-05T15:24:02-05:00

You may have never heard of the National Nuclear Security Administration, but its work is crucial to your safety—and to that of every other human being on the planet. If Elon Musk’s Department of Government Efficiency (DOGE) hasn’t yet come across the NNSA, it surely will before too long. What happens after that could be alarming.

As recently as yesterday morning, Musk made clear that DOGE will go line by line through the government’s books looking for fat targets for budget-cutting, including those that are classified—especially those that are classified. DOGE employees are bound to notice NNSA, a 1,800-person organization that sits inside the Department of Energy and burns through $20 billion every year, much of it on classified work. But as they set out to discover exactly how the money is spent, they should proceed with care. Musk’s incursions into other agencies have reportedly risked exposing sensitive information to unqualified personnel, and obstructing people’s access to lifesaving medicine. According to several nuclear-security experts and a former senior department official, taking this same approach at the NNSA could make nuclear material at home and abroad less safe.

The NNSA was created by Congress in 1999 in order to consolidate several Department of Energy functions under one bureaucratic roof: acquiring fissile material, manufacturing nuclear weapons, and preventing America’s nuclear technology from leaking. It has all manner of sensitive information on hand, including nuclear-weapon designs and the blueprints for reactors that power Navy ships and submarines. Even the Australian Navy, which has purchased some of these submarines, is not privy to their precise inner workings, James Acton, a co-director of the Nuclear Policy Program at the Carnegie Endowment for International Peace, told me.

So far, the people who work for DOGE have not wished to be slowed down by cumbersome information-security protocols. Late last week, they reportedly demanded access to a sensitive Treasury Department system that controls government payments. When the most senior civil servant at the Treasury raised security concerns, DOGE engineers were undeterred, according to The New York Times. They were happy to blast ahead while he resigned in protest.

The employees at DOGE are reportedly working seven days a week, on very little sleep. This slumber-party atmosphere isn’t a great fit for the sober and secretive world of nuclear weapons, where security lapses are hugely consequential. I spoke with three former officials and nuclear experts about what might happen if DOGE were to take a too-cavalier approach to the NNSA. None believed that Musk’s auditors would try to steal important information—although it is notable that not everyone at DOGE is a federal employee, many lack the security clearance to access the information they are seeking, and Musk had to be stopped from hiring a noncitizen. Nuclear-security lapses don’t need to be intentional to cause lasting damage. “When access to the NNSA’s sensitive systems is not granted through proper channels, they can be compromised by accident,” the former senior official at the Department of Energy, who requested anonymity to discuss internal matters, told me. “You could stumble across some incredibly sensitive things if you are coming at it sideways.”

DOGE employees might try to avoid file systems that are known to contain nuclear-weapons designs. But they could still create some risk simply by inquiring into the ways that the NNSA spends money abroad, Acton said. (Overseas expenditures have been a focus for DOGE.) The NNSA helps other governments keep highly enriched uranium secure within their own borders, and also arranges for them to ship it to the United States for safekeeping. The details of these agreements may include information about the degree to which a country’s uranium is enriched, its precise whereabouts, and the nature of the security systems that protect it—all of which are very sensitive. If one of Musk’s recruits were to access this information on their personal laptop, they could expose those secrets to hackers or spies.

[Read: The dictatorship of the engineer]

A terrorist in possession of such information could find it easier to steal material for a nuclear device, Acton said. Even the mere perception that DOGE was not minding proper security protocols could hinder the NNSA’s relationships with other countries, which are essential to its nonproliferation work. These countries may not feel like they can trust the U.S. during a security breach or other kinds of emergencies.

One nuclear-security expert with more than 10 years experience told me that he’s worried that DOGE employees will poke around in personnel records at the NNSA, as they have at other federal agencies. (The expert did not wish to be identified, because he has previously worked with the United States government and governments abroad.) As part of a larger inquiry into which employees are most productive and who gets paid what, they could potentially access the “SF-86” forms that federal employees fill out when applying for a security clearance. Those may contain information about a person’s vulnerabilities that would be useful to the hostile foreign governments that hope to recruit NNSA employees to their cause.

[Read: The growing incentive to go nuclear]

On a Monday-night conference call for concerned federal workers organized by Representative Don Beyer of Virginia, a federal contractor who works with the Energy Department asked what to do if DOGE demands access to classified nuclear data. They wouldn’t be able to complain to the inspector general. Donald Trump reportedly fired the one who oversees the Department of Energy on his fourth day in office. On the call, they were told to speak with security officials at their agency. But this is cold comfort: When DOGE employees tried to access a secure system at USAID that included personnel files, John Voorhees, that agency’s director of security, confronted them. The DOGE employees threatened to call the U.S. Marshals, and in the ensuing standoff, DOGE prevailed. Voorhees and his deputy were placed on administrative leave.

None of this is to say that the NNSA should be exempted from questions about its budget. The agency likely overspends on some things, as any bureaucracy will. But nonexperts will struggle to determine what is essential and what is excessive in its highly specialized and technical realm. Building nuclear weapons is not like making widgets. DOGE can try to root out waste, but it should take its time and avoid the break-it-to-rebuild-it approach that Musk tends to prefer. A tech-start-up mindset might be dangerous, the former official told me: “That doesn’t work with nuclear weapons.”