paywall

If Zohran Mamdani, the Democratic candidate for mayor of New York City, wins this fall’s election, he will occupy the most powerful executive position of any American socialist. At the moment, the closest contenders are two mayors in California and a county executive in Maryland. No wonder, then, that American socialists have begun to dream big.

Earlier this month, the Democratic Socialists of America, of which Mamdani is a member, held its biannual convention in Chicago, attended by 1,500 members. There, the organization pledged to “build a broad left-labor coalition” and “draft a socialist candidate” to run for president in 2028.

Why shouldn’t they? Mamdani’s primary campaign in New York showed that an appealing socialist candidate with a strong economic message could generate voter enthusiasm. Nor is Mamdani the first: Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez remain popular with the Democratic base, and just this year they brought out tens of thousands of supporters with their “Fighting Oligarchy” tour. According to a recent poll, 67 percent of Democrats and 62 percent of all Americans under 30 years old view socialism favorably. And with 85,000 members, the DSA is already the country’s biggest leftist organization, larger even than the Communist Party during its heyday in the 1940s.

But such numbers don’t add up to much political power in a country of 340 million. DSA counts only three representatives in Congress and no senators (Sanders is sympathetic but has never been a member). For the average voter, even 10 years after Sanders’s historic presidential run, American socialists are simply not a distinct, recognizable political force. And the reasons for this failure were entirely manifest at the meeting in Chicago: A significant part of the organization doesn’t share its traditional concept of an electoral path to socialism.

[Michael Powell: The mainstreaming of Zohran Mamdani]

Little about this convention suggested a mass political movement intent on winning elections and coming to power. Mamdani, AOC, and Sanders were absent, and so was their welcoming, practical political style. In fact, DSA’s national leadership has voted not to endorse AOC, and many in the organization are now actively hostile to her. Some even put forward a resolution at the convention to formally censure her for her “tacit support of Zionism,” on the grounds that she had supported the funding of defensive Iron Dome weapons for Israel, said that Israel had a right to self-defense, and “failed to support Palestinian resistance” in a media interview. (The resolution never reached a vote.) The mostly young and white crowd hardly discussed Donald Trump’s presidency (a motion that urged such discussion was voted down early on) and seemed to consist of a consortium of activists, many of them focused on single issues. Some were preoccupied with protesting the convention’s lack of a masking mandate.

Many of the resolutions passed at the convention would have been nonstarters for national politicians such as Sanders or AOC. One pledged for the DSA to be a “fighting anti-Zionist” organization that would endorse only candidates who supported the BDS (boycott, divestment, sanctions) movement against Israel. (This would disqualify Sanders.) The resolution further called for any DSA member opposing BDS or affiliated with the liberal Jewish organization J Street to be expelled, along with anybody who believes that “Israel has a right to defend itself.” Moreover, some incidents at the convention cast serious doubt on DSA’s commitment to the “democratic” part of its title. For example, the convention rejected an amendment to a resolution declaring that DSA stood “against all governments that engage in the repression of democratic rights.”

These were not the politics of DSA’s visionary founder, Michael Harrington. A Marxist who died in 1989, Harrington called for solidarity with leftist movements around the world but also staunchly opposed authoritarianism. Many in today’s DSA don’t have time for him. In response to a post on X affirming Harrington’s opposition to Stalin and Mao, DSA’s chapter in Worcester, Massachusetts, posted a video of an attendee at the convention shouting, “Fuck you, Mike Harrington!” The Worcester chapter celebrated that the organization had put “more communists in leadership,” making it the “largest org of socialists, antizionists, and commies.”

This is not rhetoric or politics that could win elections in America. But to understand the discrepancy between the politics on the DSA convention floor and that of America’s most popular socialist politicians requires a brief history. Harrington founded DSA in 1982 on the ashes of the Socialist Party of America, which had imploded a decade earlier, and the new party drew on the remnants of the youth- and student-led New Left of the 1960s. Unlike other American socialists, who would spend decades trying to establish an alternative to the duopoly of Democrats and Republicans, Harrington’s DSA was lodged within the Democratic Party and sought to build a base for the left inside of it. The strategy of the far-leftists produced little more than an alphabet soup of avowedly socialist organizations that rarely surpassed a few hundred members. DSA, too, was a tiny organization of little political account for many decades. But when American socialism finally got its lucky break in 2016, it did so because Sanders ran in the Democratic Party primary, not because socialists launched another quixotic third-party campaign. A surge of popularity for democratic socialism took DSA, within a year or two, from an organization of roughly 6,000 people, with an average age of about 67, to one with more than 30,000 members, and an average age of 33.

The infusion of new blood overwhelmed the organization’s leadership and suggested a need for some overhauling of its earlier mission. Some of the newcomers, including the young contributors and reading clubs gathered around the journal Jacobin, attempted to update Harrington’s Cold War–era socialism. But the wave of new members also included an inchoate collection of activists, and the organization swiftly became a big tent for all manner of leftist tendencies—including many that lacked any commitment to Harrington’s democratic tradition, some even holding that elections were a capitalist-state apparatus that socialists should not use to come to power.

DSA today has about two dozen internal factions (called “caucuses”), but its politics can really be divided into two broad wings. There is a mass-politics wing (grouped in the Socialist Majority and Groundwork caucuses), which seeks to elect socialists as Democrats and build a national organization that connects with the average American. Opposing it is a sectarian wing whose extremist politics have little to do with any notion of democratic socialism. The latter includes Red Star, a self-avowed “Marxist-Leninist caucus” that openly supports Hamas and emphasizes “the role of the vanguard in organizing the revolution.” Whereas the likes of Sanders have long lauded the New Deal, this group condemns that model as “extending concessions to the white working class to secure their loyalty to the capitalist state.” Similarly, it faults the Green New Deal that Sanders and AOC have championed for failing to articulate “a clear commitment to dismantling the settler-colonial and American imperialist projects.” Another caucus, Marxist Unity Group, calls for DSA “to free itself from the Democratic Party” and “fight to overthrow the Constitution,” in an effort to “destroy every institution that denies the people an authentic popular democracy, abolishing the Senate, the Electoral College, the Supreme Court, and the independent presidency.”

The differences between these two broad groups are not academic, and they have had real-life consequences. Under the pressure of the sectarian wing, the DSA refused to endorse Joe Biden or Kamala Harris in the past two presidential elections. In November 2023, the sectarians in the DSA leadership argued that a second Biden term would be no different from a second Trump term. A few months later, when the mainstream wing wanted to commit the DSA to “work to defeat Trump in the 2024 election, without endorsing the Democratic nominee,” the sectarians voted even that proposal down.

[Michael Powell: Why older socialists are quitting the DSA]

Most DSA members don’t belong to any caucuses and don’t play an active role in the organization. But since 2023, the organization’s leadership has been effectively controlled by the sectarian wing, which won a majority in that year’s convention. As anybody with experience in politics can tell you, committed sectarian activists who show up to enough meetings can capture leadership positions and convention delegates without necessarily representing the organization’s actual membership. At the 2025 convention, the mainstream wing tried to pass a resolution for the leadership to be elected on the basis of one member, one vote, as opposed to being voted in by delegates to the convention (who are, in turn, elected by DSA’s local chapters). The sectarian wing opposed and defeated the resolution. The Chicago convention elected a 27-member leadership of which the sectarian wing controls about 12 seats. The mainstream wing has about nine seats, and the rest fall somewhere in the middle. The convention also reelected the party’s two co-chairs, one belonging to Groundwork, the other to Red Star.

The two wings are able to share power this way because DSA is extremely decentralized. The organization barely exists as a nationwide project. Instead, each branch does its own thing. By far the largest branch is the one in New York City, which has more than 11,000 members and is controlled by the mainstream wing. More than 80 percent of its membership in the Bronx and Queens voted to endorse AOC. The sectarian wing tends to dominate in smaller cities where it pursues a variety of projects. Such factionalism effectively prevents DSA from adopting any unified strategy.

The problem is not new on the left. Harrington himself once complained about a “vocal, and regularly televised, fringe of confrontationists, exhibitionists, and Vietcong flag wavers who could plausibly be dismissed as freakish, or sinister, or both.” Democratic socialists who seek to run mass campaigns and attain power with elections are now encumbered by sharing an organization with “confrontationists” who hold fundamentally antidemocratic beliefs. If they wish to build a political force capable of coming to power, they must first decide who their allies are.

From The Atlantic via this RSS feed

No one doubts that our future will feature more automation than our past or present. The question is how we get from here to there, and how we do so in a way that is good for humanity.

Sometimes it seems the most direct route is to automate wherever possible, and to keep iterating until we get it right. Here’s why that would be a mistake: imperfect automation is not a first step toward perfect automation, anymore than jumping halfway across a canyon is a first step toward jumping the full distance. Recognizing that the rim is out of reach, we may find better alternatives to leaping—for example, building a bridge, hiking the trail, or driving around the perimeter. This is exactly where we are with artificial intelligence. AI is not yet ready to jump the canyon, and it probably won’t be in a meaningful sense for most of the next decade.

Rather than asking AI to hurl itself over the abyss while hoping for the best, we should instead use AI’s extraordinary and improving capabilities to build bridges. What this means in practical terms: We should insist on AI that can collaborate with, say, doctors—as well as teachers, lawyers, building contractors, and many others—instead of AI that aims to automate them out of a job.

Radiology provides an illustrative example of automation overreach. In a widely discussed study published in April 2024, researchers at MIT found that when radiologists used an AI diagnostic tool called CheXpert, the accuracy of their diagnoses declined. “Even though the AI tool in our experiment performs better than two-thirds of radiologists,” the researchers wrote, “we find that giving radiologists access to AI predictions does not, on average, lead to higher performance.” Why did this good tool produce bad results?

A proximate answer is that doctors didn’t know when to defer to the AI’s judgment and when to rely on their own expertise. When AI offered confident predictions, doctors frequently overrode those predictions with their own. When AI offered uncertain predictions, doctors frequently overrode their own better predictions with those supplied by the machine. Because the tool offered little transparency, radiologists had no way to discern when they should trust it.

A deeper problem is that this tool was designed to automate the task of diagnostic radiology: to read scans like a radiologist. But automating a radiologist’s entire diagnostic job was infeasible because CheXpert was not equipped to process the ancillary medical histories, conversations, and diagnostic data that radiologists rely on for interpreting scans. Given the differing capabilities of doctors and CheXpert, there was potential for virtuous collaboration. But CheXpert wasn’t designed for this kind of collaboration.

When experts collaborate, they communicate. If two clinicians disagree on a diagnosis, they might isolate the root of the disagreement through discussion (e.g., “You’re overlooking this.”). Or they might arrive at a third diagnosis that neither had been considering. That’s the power of collaboration, but it cannot happen with systems that aren’t built to listen. Where CheXpert’s and the radiologist’s assessments differed, the doctor was left with a binary choice: go with the software’s statistical best guess or go with her own expert judgment.

It’s one thing to automate tasks, quite another to automate whole jobs. This particular AI was designed as an automation tool, but radiologists’ full scope of work defies automation at present. A radiological AI could be built to work collaboratively with radiologists, and it’s likely that future tools will be.

Tools can be generally divided into two main buckets: In one bucket, you’ll find automation tools that function as closed systems that do their work without oversight—ATMs, dishwashers, electronic toll takers, and automatic transmissions all fall into this category. These tools replace human expertise in their designated functions, often performing those functions better, cheaper, and faster than humans can. Your car, if you have one, probably shifts gears automatically. Most new drivers today will never have to master a stick shift and clutch.

In the second bucket you’ll find collaboration tools, such as chain saws, word processors, and stethoscopes. Unlike automation tools, collaboration tools require human engagement. They are force multipliers for human capabilities, but only if the user supplies the relevant expertise. A stethoscope is unhelpful to a layperson. A chainsaw is invaluable to some, dangerous to many.

Automation and collaboration are not opposites, and are frequently packaged together. Word processors automatically perform text layout and grammar checking even as they provide a blank canvas for writers to express ideas. Even so, we can distinguish automation from collaboration functions. The transmissions in our cars are fully automatic, while their safety systems collaborate with their human operators to monitor blind spots, prevent skids, and avert impending collisions.

AI does not go neatly into either the automation bucket or the collaboration bucket. That’s because AI does both: It automates away expertise in some tasks and fruitfully collaborates with experts in others. But it can’t do both at the same time in the same task. In any given application, AI is going to automate or it’s going to collaborate, depending on how we design it and how someone chooses to use it. And the distinction matters because bad automation tools—machines that attempt but fail to fully automate a task—also make bad collaboration tools. They don’t merely fall short of their promise to replace human expertise at higher performance or lower cost, they interfere with human expertise, and sometimes undermine it.

The promise of automation is that the relevant expertise is no longer required from the human operator because the capability is now built-in. (And to be clear, automation does not always imply superior performance—consider self-checkout lines and computerized airline phone agents.) But if the human operator’s expertise must serve as a fail-safe to prevent catastrophe—guarding against edge cases or grabbing the controls if something breaks—then automation is failing to deliver on its promise. The need for a fail-safe can be intrinsic to the AI, or caused by an external failure—either way, the consequences of that failure can be grave.

The tension between automation and collaboration lies at the heart of a notorious aviation accident that occurred in June 2009. Shortly after Air France Flight 447 left Rio De Janeiro for Paris, the plane’s airspeed sensors froze over—a relatively routine, transitory instrument loss due to high-altitude icing. Unable to guide the craft without airspeed data, the autopilot automatically disengaged as it was set to do, returning control of the plane to the pilots. The MIT engineer and historian David Mindell described what happened next in his 2015 book, Our Robots, Ourselves:

When the pilots of Air France 447 were struggling to control their airplane, falling ten thousand feet per minute through a black sky, pilot David Robert exclaimed in desperation, “We lost all control of the airplane, we don’t understand anything, we’ve tried everything!” At that moment, in a tragic irony, they were actually flying a perfectly good airplane … Yet the combination of startle, confusion, at least nineteen warning and caution messages, inconsistent information, and lack of recent experience hand-flying the aircraft led the crew to enter a dangerous stall. Recovery was possible, using the old technique for unreliable airspeed—lower the pitch angle of the nose, keep the wings level, and the airplane will fly as predicted—but the crew could not make sense of the situation to see their way out of it. The accident report called it “total loss of cognitive control of the situation.”

This wrenching and ultimately fatal sequence of events puts two design failures in sharp relief. One is that the autopilot was a poor collaboration tool. It eliminated the need for human expertise during routine flying. But when expert judgment was most needed, the autopilot abruptly handed control back to the startled crew, and flooded the zone with urgent, confusing warnings. The autopilot was a great automation tool—until it wasn’t, when it offered the crew no useful support. It was designed for automation, not for collaboration.

The second failure, Mindell argued, was that the pilots were out of practice. No surprise: The autopilot was beguilingly good. Human expertise has a limited shelf life. When machines provide automation, human attention wanders and capabilities decay. This poses no problem if the automation works flawlessly or if its failure (perhaps due to something as mundane as a power outage) doesn’t create a real-time emergency requiring human intervention. But if human experts are the last fail-safe against catastrophic failure of an automated system—as is currently true in aviation—then we need to vigilantly ensure that humans attain and maintain expertise.

Modern airplanes have another cockpit navigation aid, one that is less well known than the autopilot: the heads-up display. The HUD is a pure collaboration tool, a transparent LCD screen that superimposes flight data in the pilot’s line of sight. It does not even pretend to fly the aircraft, but it assists the pilot by visually integrating everything that the flight computer digests about the plane’s direction, pitch, power, and airspeed into a single graphic called the flight-path vector. Absent a HUD, a pilot must read multiple flight instruments to intuitively stitch this picture together. The HUD is akin to the navigation app on your smartphone—if that app also had night vision, speed sensors, and intimate knowledge of your car’s engine and brakes.

The HUD is still a piece of complex software, meaning it can fail. But because it is built to collaborate and not to automate, the pilot continually maintains and gains expertise while flying with it—which, to be clear, is typically not the whole flight, but in crucial moments such as low-visibility takeoff, approach, and landing. If the HUD reboots or locks up during a landing, there is no abrupt handoff; the pilot already has hands on the control yoke for the entire time. Despite the fact that HUDs offer less automation than automatic landing systems, airlines have discovered that their planes suffer fewer costly tail strikes and tire blowouts when pilots use HUDs rather than auto-landers. Perhaps for this reason, HUDs are integrated into newer commercial aircraft.

Collaboration is not intrinsically better than automation. It would be ridiculous to collaborate with your car’s transmission or to pilot your office elevator from floor to floor. But in some domains, occupations, or tasks where full automation is not currently achievable, where human expertise remains indispensable or a necessary fail-safe, tools should be designed to collaborate—to amplify human expertise, not to keep it on ice until the last possible moment.

One thing that our tools have not historically done for us is make expert decisions. Expert decisions are high-stakes, one-off choices where the single right answer is not clear—often not knowable—but the quality of the decision matters. There is no single best way, for example, to care for a cancer patient, write a legal brief, remodel a kitchen, or develop a lesson plan. But the skill, judgment, and ingenuity of human decision making determines outcomes in many of these tasks, sometimes dramatically so. Making the right call means exercising expert judgment, which means more than just following the rules. Expert judgment is needed precisely where the rules are not enough, where creativity, ingenuity, and educated guesses are essential.

But we should not be too impressed by expertise: Even the best experts are fallible, inconsistent, and expensive. Patients receiving surgery on Fridays fare worse than those treated on other days of the week, and standardized test takers are more likely to flub equally easy questions if they appear later on a test. Of course, most experts are far from the best in their fields. And experts of all skill levels may be unevenly distributed or simply unavailable—a shortage that is more acute in less affluent communities and lower-income countries.

Expertise is also slow and costly to acquire, requiring immersion, mentoring, and tons of practice. Medical doctors—radiologists included—spend at least four years apprenticing as residents; electricians spend four years as apprentices and then another couple as journeymen, before certifying as master electricians; law-school grads start as junior partners, and new Ph.D.s begin as assistant professors; pilots must log at least 1,500 hours of flight before they can apply for an Airline Transport Pilot license.

The inescapable fact that human expertise is scarce, imperfect, and perishable makes the advent of ubiquitous AI an unprecedented opportunity. AI is the first machine humanity has devised that can make high-stakes, one-off expert decisions at scale—in diagnosing patients, developing lesson plans, redesigning kitchens. AI’s capabilities in this regard, while not perfect, have consistently been improving year by year.

What makes AI such a potent collaborator is that it is not like us. A modern AI system can ingest thousands of medical journals, millions of legal filings, or decades of maintenance logs. This allows it to surface patterns and keep up with the latest developments in health care, law, or vehicle maintenance that would elude most humans. It offers breadth of experience that crosses domains and the capacity to recognize subtle patterns, interpolate among facts, and make new predictions. For example, Google DeepMind’s AlphaFold AI overcame a central challenge in structural biology that has confounded scientists for decades: predicting the folding labyrinthine structure of proteins. This accomplishment is so significant that its designers, Demis Hassabis and John Jumper, colleagues of one of us, were awarded the Nobel Prize in Chemistry last year for their work.

The question is not whether AI can do things that experts cannot do on their own—it can. Yet expert humans often bring something that today’s AI models cannot: situational context, tacit knowledge, ethical intuition, emotional intelligence, and the ability to weigh consequences that fall outside the data. Putting the two together typically amplifies human expertise: Oncologists can ask a model to flag every recorded case of a rare mutation and then apply clinical judgment to design a bespoke treatment; a software architect can have the model retrieve dozens of edge-case vulnerabilities and then decide which security patch best fits the company’s needs. The value is not in substituting one expert for another, or in outsourcing fully to the machine, or indeed in presuming the human expertise will always be superior, but in leveraging human and rapidly-evolving machine capabilities to achieve best results.

As AI’s facility in expert judgment becomes more reliable, capable, and accessible in the years ahead, it will emerge as a near-ubiquitous presence in our lives. Using it well will require knowing when to automate versus when to collaborate. This is not necessarily a binary choice, and the boundaries between human expertise and AI’s capabilities for expert judgment will continually evolve as AI’s capabilities advance. AI already collaborates with human drivers today, provides autonomous taxi services in some cities, and may eventually relieve us of the burden and risk of driving altogether—so that the driver’s license can go the way of the manual transmission. Although collaboration is not intrinsically better than automation, premature or excess automation—that is, automation that takes on entire jobs when it’s ready for only a subset of job tasks—is generally worse than collaboration.

The temptation toward excess automation has always been with us. In 1984, General Motors opened its “factory of the future” in Saginaw, Michigan. President Ronald Reagan delivered the dedication speech. The vision, as MIT’s Ben Armstrong and Julie Shaw wrote in Harvard Business Review in 2023, was that robots would be “so effective that people would be scarce—it wouldn’t even be necessary to turn on the lights.” But things did not go as planned. The robots “struggled to distinguish one car model from another: They tried to affix Buick bumpers to Cadillacs, and vice versa,” Armstrong and Shaw wrote. “The robots were bad painters, too; they spray-painted one another rather than the cars coming down the line. GM shut the Saginaw plant in 1992.”

There has been much progress in robotics since this time, but the advent of AI invites automation hubris to an unprecedented degree. Starting from the premise that AI has already attained superhuman capabilities, it is tempting to think that it must be able to do everything that experts do, minus the experts. Many people have therefore adopted an automation mindset, in their desire either to evangelize AI or to warn against it. To them, the future goes like this: AI replicates expert capabilities, overtakes the experts, and finally replaces them altogether. Rather than performing valuable tasks expertly, AI makes experts irrelevant.

Research on people’s use of AI makes the downsides of this automation mindset ever more apparent. For example, while experts use chatbots as collaboration tools—riffing on ideas, clarifying intuitions—novices often treat them mistakenly as automation tools, oracles that speak from a bottomless well of knowledge. That becomes a problem when an AI chatbot confidently provides information that is misleading, speculative, or simply false. Because current AIs don’t understand what they don’t understand, those lacking the expertise to identify flawed reasoning and outright errors may be led astray.

The seduction of cognitive automation helps explain a worrying pattern: AI tools can boost the productivity of experts but may also actively mislead novices in expertise-heavy fields such as legal services. Novices struggle to spot inaccuracies and lack efficient methods for validating AI outputs. And methodically fact-checking every AI suggestion can negate any time savings.

Beyond the risk of errors, there is some early evidence that overreliance on AI can impede the development of critical thinking, or inhibit learning. Studies suggest a negative correlation between frequent AI use and critical-thinking skills, likely due to increased “cognitive offloading”—letting the AI do the thinking. In high-stakes environments, this tendency toward overreliance is particularly dangerous: Users may accept incorrect AI suggestions, especially if delivered with apparent confidence.

The rise of highly capable assistive AI tools also risks disrupting traditional pathways for expertise development when it’s still clearly needed now, and will be in the foreseeable future. When AI systems can perform tasks previously assigned to research assistants, surgical residents, and pilots, the opportunities for apprenticeship and learning-by-doing disappear. This threatens the future talent pipeline, as most occupations rely on experiential learning—like those radiology residents discussed above.

Early field evidence hints at the value of getting this right. In a PNAS study published earlier this year and covering 2,133 “mystery” medical cases, researchers ran three head-to-head trials: doctors diagnosing on their own, five leading AI models diagnosing on their own, and then doctors reviewing the AI suggestions before giving a final answer. That human-plus-AI pair proved most accurate, correct on roughly 85 percent more cases than physicians working solo and 15 to 20 percent more than an AI alone. The gain came from complementary strengths: When the model missed a clue, the clinician usually spotted it, and when the clinician slipped, the model filled the gap. The researchers engineered human-AI complementarity into the design of the trials, and saw results. As these tools evolve, we believe they will surely take on autonomous diagnostic tasks, such as triaging patients and ordering further testing—and may indeed do better over time on their own, as some early studies suggest.

Or, consider an example with which one of us is closely familiar: Google’s Articulate Medical Intelligence Explorer (AMIE) is an AI system built to assist physicians. AMIE conducts multi-turn chats that mirror a real primary-care visit: It asks follow-up questions when it is unsure, explains its reasoning, and adjusts its line of inquiry as new information emerges. In a blinded study recently published in Nature, specialist physicians compared the performance of a primary-care doctor working alone with that of a doctor who collaborated with AMIE. The doctor who used AMIE ranked higher on 30 of 32 clinical-communication and diagnostic axes, including empathy and clarity of explanations.

By exposing its reasoning, highlighting uncertainty, and grounding advice in trusted sources, AMIE pulls the user into an active problem-solving loop instead of handing down answers from on high. Doctors can potentially interrogate and correct it in real time, reinforcing (rather than eroding) their own diagnostic skills. These results are preliminary: AMIE is still a research prototype and not a drop-in replacement. But its design principles suggest a path toward meaningful human collaboration with AI.

Full automation is much harder than collaboration. To be useful, an automation tool must deliver near flawless performance almost all of the time. You wouldn’t tolerate an automatic transmission that sporadically failed to shift gears, an elevator that regularly got stuck between floors, or an electronic tollbooth that occasionally overcharged you by $10,000.

By contrast, a collaboration tool doesn’t need to be anywhere close to infallible to be useful. A doctor with a stethoscope can better understand a patient than the same doctor without one; a contractor can pitch a squarer house frame with a laser level than by line of sight. These tools don’t need to work flawlessly, because they don’t promise to replace the expertise of their user. They make experts better at what they do—and extend their expertise to places it couldn’t go unassisted.

Designing for collaboration means designing for complementarity. AI’s comparative advantages (near limitless learning capacity, rapid inference, round-the-clock availability) should slot into the gaps where human experts tend to struggle: remembering every precedent, canvassing every edge case, or drawing connections across disciplines. And at the same time, interface design must leave space for distinctly human strengths: contextual nuance, moral reasoning, creativity, and a broad grasp of how accomplishing specific tasks achieves broader goals.

Both AI skeptics and AI evangelists agree that AI will prove a transformative technology–-indeed, this transformation is already under way. The right question then is not whether but how we should use AI. Should we go all in on automation? Should we build collaborative AI that learns from our choices, informs our decisions, and partners with us to drive better results? The correct answer, of course, is both. Getting this balance right across capabilities is a formidable and ever-evolving challenge. Fortunately, the principles and techniques for using AI collaboratively are now emerging. We have a canyon to cross. We should choose our routes wisely.

From The Atlantic via this RSS feed

Editor’s Note: Washington Week With The Atlantic is a partnership between NewsHour Productions, WETA, and The Atlantic airing every Friday on PBS stations nationwide. Check your local listings, watch full episodes here, or listen to the weekly podcast here.

This week, Director of National Intelligence Tulsi Gabbard removed the security clearances from 37 former and current intelligence officials, including those with expertise on Russian election interference. Meanwhile, former National Security Adviser John Bolton’s home was raided. Panelists on Washington Week With The Atlantic joined last night to discuss the potential implications of the Trump administration’s actions on national security, and more.

“Tulsi Gabbard came in early on talking about trying to de-politicize the intelligence community,” Matt Viser, the White House bureau chief at The Washington Post, said last night. But her latest decisions, among others, “gives the sense of political retribution.”

Joining the editor in chief of The Atlantic, Jeffrey Goldberg, to discuss this and more: Laura Barron-Lopez, a White House correspondent at MSBNC; Jonathan Karl, the chief Washington correspondent at ABC News; Michael Scherer, a staff writer at The Atlantic; and Viser, the White House bureau chief at The Washington Post.

Watch the full episode here.

From The Atlantic via this RSS feed

Welcome back to the Abstract! Here are the studies this week that ruled the roost, warmed the soul, and departed for intergalactic frontiers.

It will be a real creature feature this week. First, we will return to the realm of bats and discover that it is, in fact, still awesome. Then: poops from above; poops from the past; a very special bonobo; and last, why some dead stars are leaving the Milky Way in a hurry.

Bat hugs > Bear hugs

Tietge, Marisa et al. “Cooperative behaviors and social interactions in the carnivorous bat Vampyrum spectrum.” PLOS One.

Welcome to The Real World: Bat Roost. Scientists installed a camera into a tree hollow in Guanacaste, Costa Rica to film a tight-knit family of four spectral bats (Vampyrum spectrum) over the course of several months. The results revealed many never-before-seen behaviors including bats hugging, playing with cockroaches, and even breaking the fourth wall.

“We provide the first comprehensive account of prey provision and other social behaviors in the spectral bat V. spectrum,” said researchers led by Marisa Tietge of Humboldt University in Berlin. “By conducting extensive video recordings in their roost, we aimed to document and analyze key behaviors.”

Spectral bats are the biggest bats in the New World, with wingspans that can exceed three feet. They are carnivorous—feasting on rodents, birds, and even other species of bat—and they mate in monogamous pairs, which is unusual for mammals. But while huge flesh-eating bats sound scary, the new study revealed that these predators have a soft side.

For example, the footage captured a “greeting” ritual that included “a hugging-like interaction between a bat already in the roost and a newly arrived bat,” according to the study.

“The resident bat may actively approach or greet the newcomer as it reaches close proximity in the main roosting area,” the team said. “The greeting behavior is comparable to the initiation to social roosting, where at least one bat wraps its wings around the other, establishing a ball-like formation for several seconds. This behavior is often accompanied by social vocalizations.”

There’s nothing like coming home after a graveyard shift to a warm welcome in a fuzzy ball-like formation. In keeping with their gregarious nature, the footage also showed that the bats are very generous with sharing prey, with only a single instance of a “tug-of-war” breaking out over dinner.

“Prey provision was a clearly cooperative social behavior wherein a bat successfully captured prey, brought it to the roost where group members were present, and willingly transferred the prey to another bat,” the researchers said. “Audible chewing noises are a distinctive feature of this process.”

Loud chewers in any other context are profoundly irritating, but these bats get a pass because it’s kind of hard to be quiet while crunching through mouse bones perched upside-down.

In addition to all the hugging and prey-sharing, the bats were also observed playing together by chasing cockroaches or, in one case, messing with the camera by altering its position. I can’t wait for the next season!

In other news…

Skyward scat

Uesaka, Leo and Sato, Katsufumi. “Periodic excretion patterns of seabirds in flight.” Current Biology.

Speaking of putting cameras in weird places, why not strap them to the bellies of seabirds? Scientists went ahead and did this, ostensibly to examine the flight dynamics of streaked shearwaters, which are Pacific seabirds. But the tight focus on the bird-bums produced a different revelation: Shearwaters almost exclusively poop while on the wing.

“A total of 195 excretions were observed from 35.9 hours of video data obtained from 15 streaked shearwaters,” said authors Leo Uesaka and Katsufumi Sato of the University of Tokyo. “Excretion immediately after takeoff was frequent, with 50 percent of the 82 first excretion events during the flying periods occurring within 30 seconds after take-off and 36.6 percent within 10 seconds.”

“Occasionally, birds took off, excreted, and returned to the water within a minute; these take-offs are speculated to be only for excretion,” the team continued. “These results strongly suggest that streaked shearwaters intentionally avoid excretion while floating on the sea surface.”

This preference for midair relief might allow seabirds to lighten their load, prevent backward contamination, and avoid predators that sniff out excrement. Whatever the reason, these aerial droppings provide nutrients to ocean ecosystems, so bombs away.

Please clean up after your 9,000-year-old dog

Slepchenko, S.M. et al. “Early history of parasitic diseases in northern dogs revealed by dog paleofeces from the 9000-year-old frozen Zhokhov site in the New Siberian Islands of East Siberian Arctic.” Journal of Archaeological Science.

Hold onto your butts, because we’re not done with scatological science yet. A study this week stepped into some very ancient dog doo recovered from a frozen site on Siberia’s Zhokhov Island, which was inhabited by Arctic peoples 9,000 years ago.

By analyzing the “paleofeces,” scientists were able to reconstruct the diet of these canine companions, which were bred in part as sled dogs. The results provide the first evidence of parasites in Arctic dogs of this period, suggesting that the dogs were fed raw fish, reindeer, and polar bear.

“The high infection rate in dogs with diphyllobothriasis indicates a significant role of fishing in the economic activities of Zhokhov inhabitants, despite the small amount of direct archaeological evidence for this activity,” said researchers led by S.M. Slepchenko of Tyumen Scientific Center. “The presence of Taeniidae eggs indicates that dogs were fed reindeer meat.”

The team also noted that after excavation, the excrement samples were “packaged entirely in a separate hermetically sealed plastic bag and labeled.” It seems even prehistoric dog poop ends up in plastic bags.

Kanzi the unforgettable bonobo

Carvajal, Luz and Krupenye, Christopher. “Mental representation of the locations and identities of multiple hidden agents or objects by a bonobo.” Proceedings of the Royal Society B.

Playing hide-and-seek with bonobos is just plain fun, but it also doubles as a handy experiment for testing whether these apes—our closest living relatives—can track the whereabouts of people, even when they are out of sight.

Kanzi, a bonobo known for tool use and language skills, participated in experiments in which his caretakers hid behind screens. He was asked to identify them from pictures or voices and succeeded more than half the time, above chance (here’s a video of the experiment).

”Kanzi presented a unique and powerful opportunity to address our question in a much more straightforward way than would be possible with almost any other ape in the world,” said authors Luz Carvajal and Christopher Krupenye of Johns Hopkins University. “He exhibited not only strong engagement with cognitive tasks but also rich forms of communication with humans—including pointing, use of symbols, and response to spoken English.”

Kanzi was also a gamer who played Pac-Man and Minecraft. Image: William H. Calvin, PhD -

Sadly, this was one of Kanzi’s last amazing feats, as he died in March at the age of 44 in his long-time home at the Ape Initiative in Des Moines, Iowa. But as revealed by this posthumous study, Kanzi’s legacy as a cognitive bridge between apes lives on. RIP to a real one.

Zero to 4.5 million mph in a millisecond

Glanz, Hila and Perets Hagai B. et al. “The origin of hypervelocity white dwarfs in the merger disruption of He–C–O white dwarfs.” Nature Astronomy.

We will close with dead stars that are careening out of the galaxy at incomprehensible speeds. These objects, known as hypervelocity white dwarfs, are corpses of stars similar in scale to the Sun, but it remains unclear why some of them fully yeet themselves into intergalactic space.

“Hypervelocity white dwarfs (HVWDs) are stellar remnants moving at speeds that exceed the Milky Way’s escape velocity,” said researchers co-led by Hila Glanz and Hagai B. Perets of the Technion–Israel Institute of Technology. “The origins of the fastest HVWDs are enigmatic, with proposed formation scenarios struggling to explain both their extreme velocities and observed properties.”

The team modeled a possible solution that involves special white dwarfs with dense carbon-oxygen cores and outer layers of helium, known as hybrid helium-carbon-oxygen (HeCO) white dwarfs. When two He-CO white dwarfs merge, it may trigger a “double-detonation explosion” that launches one of the objects to speeds of about 4.5 million miles per hour.

“We have demonstrated that the merger of two HeCO white dwarfs can produce HVWDs with properties consistent with observations” which “provides a compelling explanation for the origin of the fastest HVWDs and sheds new light on the diversity of explosive transients in the Universe,” the researchers concluded.

With that, may you sail at hypervelocity speeds out of this galaxy and into the weekend.

Thanks for reading! See you next week.

From 404 Media via this RSS feed

This is an edition of The Wonder Reader, a newsletter in which our editors recommend a set of stories to spark your curiosity and fill you with delight. Sign up here to get it every Saturday morning.

College athletics were once casual and fun, more like a club sport than a serious endeavor. But “over the past 75 years, NCAA sports has become ever more professionalized,” Marc Novicoff wrote recently. “Football and men’s basketball began to generate eye-watering sums of money, incentivizing colleges to invest more resources in them.”

Now, recent court cases have allowed athletes to get paid by advertisers, fans, and their schools. As college athletes’ status changes, both their careers and the experience of college-level sports are starting to look different. Today’s reading list explores the meaning of school sports.

On College Sports

The End of Niche College Sports

By Marc Novicoff

Letting schools pay revenue-generating athletes is long overdue. If that means letting squash and water polo die, so be it.

Read the article.

College Sports Are Affirmative Action for Rich White Students

By Saahil Desai

Athletes are often held to a lower standard by admissions officers, and in the Ivy League, 65 percent of players are white. (From 2018)

Read the article.

The Logical End Point of College Sports

By Marc Novicoff

If players are workers, schools will have to pay them.

Read the article.

Still Curious?

Meritocracy is killing high-school sports: Athletics are supposed to be great equalizers in American life. But they’re being hijacked by the wealthy, Derek Thompson wrote in 2019.Do sports matter?: In 2022, readers weighed in on the role of athletics in today’s society—and if they should have one at all.

Other Diversions

What we gain when we stop caringA tale of sex and intrigue in imperial Kyoto Alexandra Pertri: Donald Trump’s guide to museums

P.S.

Courtesy of Nancy Farese

I recently asked readers to share a photo of something that sparks their sense of awe in the world. Nancy Farese sent this photo “of my cold dip buddies who brave the icy temperatures of the SF Bay at 7am on Sunday mornings.” She adds: “It's a time of fellowship (we circle up and read Mary Oliver's Why I Wake Early) and courage, the glossy heads of seals, the clatter of gulls, and the cloak of fog rolling back under the Golden Gate Bridge as the sun is rising.”

I’ll continue to feature your responses in the coming weeks.

— Isabel

From The Atlantic via this RSS feed

Practicing science in the United States has become more politically fraught in the past seven months than it has ever been in this country’s history. As the Trump administration has fired vaccine advisers, terminated research grants in droves, denied the existence of gender, and accused federal scientists of corruption while publicly denigrating their work, the nation’s leaders have shown that they believe American science should be done only on their terms.

As of late, some in the scientific community have been pushing back, organizing marches and rallies, publicly criticizing government reports and agency priorities, and quitting their jobs at federal agencies. Professional medical societies have banded together to sue the Department of Health and Human Services over Secretary Robert F. Kennedy Jr.’s unfounded restrictions of COVID vaccines and dismissal of vaccine experts. Academic scientists have done the same, to fight for grant funding. Researchers are convening extragovernmental panels to evaluate evidence on vaccines; the American Academy of Pediatrics has published vaccine recommendations that deviate from the CDC’s, and several states in New England are mulling doing the same. This week, for the second time, hundreds of HHS officials have signed a public letter criticizing the department’s leaders for interfering with the integrity of their work.

And yet, these counterattacks may be ensnaring scientists in a catch-22. Their goal is to defend their work from political interference. “If scientists don’t ever speak up, then the court of public opinion is lost,” one university dean, who requested anonymity to avoid financial retaliation against their school from the federal government, told me: Americans would have little reason to question the government’s actions. But in retaliating, scientists also run the risk of advancing the narrative they want to fight—that science in the U.S. is a political endeavor, and that the academic status quo has been tainted by an overly liberal view of reality. “When you face a partisan attack, it’s extremely hard to respond in a way that doesn’t look partisan,” Alexander Furnas, a science-policy expert at Northwestern University, told me. “It’s a bit of a trap.”

Many scientists prefer to view their work as largely severed from politics. But in practice, politicians control how science is funded and how its findings are codified into policy. Some science has also been actively coded as partisan: The existence of climate change has been publicly questioned by conservative groups; since the early days of COVID vaccines, skepticism of them has split along party lines. And studies show that trust in the scientific community has been eroding among conservatives since the 1970s. Still, for decades, science in the U.S. has enjoyed bipartisan support. Furnas’s unpublished research, for instance, has found that over the past 40 years, Republicans have appropriated more money to science than Democrats.

But if any previous politicizations of science were matchsticks tossed onto embers, the Trump administration “has been pouring gasoline,” Azim Shariff, a social psychologist at the University of British Columbia, who has studied how the politicization of science influences trust in it, told me. Many of the administration’s assaults have been openly political—its leaders have repeatedly criticized American research as riddled with problematic ideologies, and claimed that the Biden administration manipulated science for its own purposes. And it has treated academic centers of science as threats that must be forcibly dismantled. “There is virtually no part of science that is not seen as belonging to one side, particularly the Democrats,” the university dean told me. “Science in general has been cast as being the work of one party, while those of another party destroy the system as it exists.” (HHS and the White House did not return requests for comment.)

Government scientists in particular have usually stayed out of the political fray. The federal workforce is largely made up of rule followers, Anna Yousaf, a scientist in CDC’s respiratory-virus division who signed her name publicly to this week’s HHS letter, told me. (She and other federal employees I spoke with emphasized that they were talking in their personal capacity, rather than on behalf of their agency.) “In terms of feeling comfortable about this? I don’t,” Yousaf said. But now these scientists’ livelihoods are on the line, as well the scientific principles they’ve dedicated their careers to.

And many fear for their personal safety. Earlier this month, a man who had expressed “discontent with the COVID-19 vaccinations” fired hundreds of rounds at the CDC’s headquarters, killing a police officer. The shooting, and the administration’s muted response to it, was a major motivation for Fiona Havers, who recently quit her job at the CDC in protest of Kennedy’s actions, to sign her name to the letter. Kennedy’s inflammatory accusations about public-health officials—including calling the CDC “a cesspool of corruption”—have “endangered the lives of my friends and former colleagues,” she told me. (Kennedy’s earliest response, a post on X, came the day after the shooting; two days later, HHS released the administration’s only official statement to date. Neither acknowledged the role that misinformation about COVID vaccines may have played, and hours after HHS’s statement, Kennedy publicly criticized the CDC’s pandemic response, arguing that the government said “things that are not always true” to persuade the public to get vaccinated.)

Many of the scientists I spoke with for this story insisted that they didn’t feel their actions were political—and expressed concern over them being perceived as such. Although they were fighting back against the government, they told me, their intentions are to advocate for evidence. That line feels especially important to hold, they said, as Kennedy and other political leaders repeatedly flaunt their disregard for facts and scientific consensus. “We have not made this political,” Susan Kressly, the president of the American Academy of Pediatrics—which has sued HHS, boycotted meetings of its vaccine advisory committee, and continued to recommend COVID-19 vaccines for populations that the CDC does not—told me. “It is the politicians doing that.” Georges Benjamin, the executive director of the American Public Health Association, one of the professional societies that has sued HHS, told me that he felt similarly. “People tend to think of us as very much left-leaning,” Benjamin said, but the APHA, like the AAP, identifies as nonpartisan. He and Kressly each pointed out that their society has criticized the government during both Democratic and Republican administrations. For example, both groups were among the organizations that, in 2024, called out the Biden administration for delaying prohibitions on menthol cigarettes and flavored cigars.

In the past, scientists have successfully engaged in advocacy without polarizing public perceptions of them and their work. And enough Americans object to the Trump administration’s campaign against science that Floyd Zhang, an economist who has studied public attitudes toward science, told me he could see trust in researchers increasing now. His research has shown that engaging politics can hurt scientists: In 2020, after the scientific journal Nature endorsed Joe Biden for president, Trump supporters who were told about the endorsement lost trust in the journal—and in scientists in general. Researchers, he said, seemed to be speaking out of turn—Who are you, telling me how to vote? But he thinks what’s happening in 2025 may play out differently. Scientists’ advocacy—for themselves, their institutions, and scientific principles—should look like scientists staying in their lane, and fighting on behalf of science.

Still, some scientists are behaving more like political activists and politicians. The writers of the HHS letters understand that defending their idea of the department requires political allies: Ian Morgan, a postdoctoral fellow at the National Institutes of Health and one of the signers of both letters, told me that a main goal of the public outcry is to stir up further congressional support. A social-media account run by anonymous NIH officials explicitly calls out the “rightwing billionaires” who are trying to corrupt their agency. And scientists and physicians have cited the Trump administration’s actions as motivation in their run for Democratic congressional seats.

Their choice of party is not just a protest against this administration. Scientists, as a group, lean more Democratic and less Republican than the rest of the public, a trend that seems to have intensified in recent decades. Pediatrics—the subgroup of medicine that communicates most regularly with families about vaccines—is among the most left-leaning medical specialties.

Already, public opinion on the Trump administration’s siege on science divides along party lines. An April poll from the health nonprofit KFF showed that a majority of Republicans supported Trump’s massive cuts to staff and spending at federal health agencies, whereas nearly all Democrats opposed them. (Another, more recent survey, from the Civic Health and Institutions Project, noted more muted enthusiasm from Republicans—but still found that more Republicans approved of Trump’s attacks on science than did not.) More Republicans than Democrats support slashing funding to universities, too. And 41 percent of Republicans say that HHS’s recent changes to vaccine policy will make people safer, compared with just 4 percent of Democrats.

Whatever the scientists’ intentions, their actions may inadvertently bolster the Trump administration’s case that scientists represent a particular liberal worldview. Shariff, the social psychologist, has found in his research that—even when politicization aligns with their own beliefs—“people don’t like to see their science politicized,” he told me. “They lose trust in it.” That decline in trust, Shariff predicts, will concentrate among those on the right, who “will see science as more politicized than they did before,” he said, “because it’s taking a side.”

If that happens, the administration could leverage the validation of public opinion as permission to escalate. Trump and his appointees have loudly asserted that their vision for science in America is the correct one, representing truth rather than politics. In their view, the problem originated with the scientists who allowed ideology to infiltrate their thinking, fell prey to the distortive influence of industry, and discouraged the public from doing “your own research.” They seem ready, too, to blame scientists for the ongoing fracas. In July, NIH Director Jay Bhattacharya sat down with some members of his staff to discuss the letter they had signed, calling for a restoration of the agency’s scientific integrity. After a fairly cordial meeting, Bhattacharya’s staff invited him to join them at a pro-NIH rally—perhaps even speak, Sarah Kobrin, a branch chief at the National Cancer Institute who attended the meeting, told me. “That appeared to anger him,” Kobrin said. Bhattacharya declined and stood to leave, adding, “I’m disappointed that you are politicizing this.”

From The Atlantic via this RSS feed

Both batches of french fries that my family had for dinner were from the frozen-food aisle. They appeared nearly identical when cooked, one batch faintly darker than the other. And we all noticed: Those bronzed fries were exceptionally tasty. My toddler devoured a small mountain of them. They left a meatiness on my tongue, as if I’d eaten them alongside a steak. After my husband unblinded the taste test, I realized that, in a way, I had. The paler fries had been cooked in avocado oil, and their more delicious counterparts in beef tallow. Damn, I thought. The MAHA fries are amazing!

They weren’t, of course, actually produced by the Make America Healthy Again campaign; both bags were from Jesse and Ben’s, a frozen-french-fry brand whose tallow fries predate Robert F. Kennedy Jr.’s tenure as secretary of Health and Human Services. Jesse and Ben’s, like many food companies, had already released so-called clean-label products, which cater to long-standing wellness trends such as avoiding artificial ingredients and added sugar—trends that overlap considerably with the MAHA approach to food.

Now companies are capitalizing on some of Kennedy’s favored dietary principles—including his assertion, which is refuted by most nutrition experts, that beef tallow is a healthy substitute for seed oils—by further overhauling the branding and recipes of their products. Unfortunately, MAHA-washing groceries in this way won’t make Americans any healthier. It might even change our diets for the worse.

Many product labels and ad campaigns decry ingredients on Kennedy’s hit list—besides seed oils, it also includes high-fructose corn syrup and artificial food dyes and flavors—and showcase those he deems healthy. This summer, Sam’s Club started selling beef-tallow-fried chicken strips. A brand of seed-oil-free instant ramen launched in August and is available at Kroger. This spring, PepsiCo relaunched its “Simply” line, which sells versions of snacks such as Cheetos and Doritos that are made without artificial flavorings and dyes; it later announced plans to extend the line with new products. A company spokesperson told me in an email that Lays and Tostitos will have no artificial colors or flavors by the end of the year. PepsiCo is investing in products without artificial dyes and flavorings “to make it easier for everyone to find snacks and drinks they feel good about,” the spokesperson told me. “The Make America Healthy Again movement has sparked important dialogue, and we share the belief that the food system should evolve—responsibly, at scale, and grounded in science and consumer trust.” Meanwhile, Coca-Cola announced that it would sweeten its sodas with cane sugar instead of high-fructose corn syrup. President Donald Trump, who said he had previously discussed the change with the company, thanked its leaders; Kennedy subsequently thanked Trump.

Of course, fried chicken, instant ramen, soda, and chips share a certain inherent junkiness. Even without their shocking-orange hue, Cheetos are mostly empty vectors for salt and fat. A 12-ounce bottle of Mexican Coke still contains more than three-quarters of the added sugar that the FDA says an adult should limit themselves to in a day. MAHA-washing therefore “misses the bigger picture of the food landscape,” which is characterized by heavily processed food, fast food, and sugary drinks, Marie Bragg, a population-health professor at New York University, told me.

These reformulations may have some benefits; as my colleague Nick Florko has written, artificial food dyes in particular are both unnecessary and probably not great for health. But at best, the changes championed by the MAHA movement will likely yield marginal health improvements, Alyssa J. Moran, a director at the University of Pennsylvania’s food-policy laboratory, told me. Research has long shown that the most harmful elements of junk food are high levels of salt, saturated fat, and sugar, combined with minimal fiber and nutrients—not fructose, seed oils, or trace amounts of additives. Despite widespread concern resulting from studies linking high-fructose corn syrup to obesity in the 2000s, the evidence that it is less healthy than other forms of sugar is weak. Seed oils have repeatedly been shown to be not only safe to consume, but healthier than animal-based fats such as butter and beef tallow, which are rich in saturated fat and are linked to higher risk of heart disease. As I read the nutrition labels of my frozen fries, my heart spasmed: The beef-tallow version contains seven times more saturated fat than the avocado-oil kind.

[Read: America stopped cooking with tallow for a reason]

Unfortunately, Americans have proved themselves to be suckers for packaging that conveys a food’s healthiness, Bragg said. Shoppers are willing to pay more for food labeled “all natural” and prefer produce marked as “pesticide-free.” One study that Moran co-authored found that parents are more likely to give their kids sugary drinks labeled with images of fresh fruit than similar products without those images. People tend to falsely believe that Oreos labeled “organic” have fewer calories than their conventional counterparts, and that the cookies can be eaten more frequently, even if they are pointed to labels showing that both options are nutritionally identical. They are also more likely to forgo exercise if they choose an organic dessert over a conventional one. All of this bodes poorly for American shoppers, who seem likely to eat more of the MAHA-washed junk foods that will still contain just as much salt, saturated fat, and sugar.

These issues do not concern food companies, whose primary mission is, of course, to sell food. Jesse Konig, one of Jesse and Ben’s co-founders, told me that the company was pursuing taste, not health, when it started selling tallow fries, in 2024. “We’re restaurant guys, not doctors,” he said. The labels on my packages of Jesse and Ben’s fries, however, noted that the company doesn’t use conventional seed oils, because they “leave you feeling icky and inflamed,” referencing a common health claim made by seed-oil critics.

Other companies are more outspoken about changing their products for the purpose of health. Mike’s Mighty Good describes its seed-oil-free ramen as “wholesome,” and similar instant-ramen products as “low-quality junk food.” Real Good Foods launched its tallow-fried chicken because customers kept asking for a “seed-oil-free solution,” Rikki Ingram, the company’s chief marketing officer, told me. Compared with conventional products, she said, the brand’s tallow-fried chicken offers health advantages unrelated to its lack of seed oils: more protein, fewer carbohydrates, and no highly processed flour.

[Read: America is done pretending about meat]

Changes such as these make good business sense. A company that agrees to, say, phase out food dyes from soft drinks improves its public image. It also courts a relatively affluent audience, says Trey Malone, an agricultural economist at Purdue University. MAHA-washed foods are likely to be more expensive, in part because reformulating products is costly; companies aren’t trying to market those goods to people already struggling to afford conventional food. Mike’s Mighty Good seed-oil-free instant ramen costs more than $3 a cup on its website; its conventional counterparts can be 99 cents or less apiece. At Walmart, a bag of Simply Lays costs nearly three times as much as regular Lays. The rise of MAHA has been good for Jesse and Ben’s business, Konig told me. Both the avocado-oil fries—which tick MAHA’s seed-oil-free criterion—and the beef-tallow version have been hits with customers, but recent public discussion about beef tallow especially has “generated a lot of curiosity,” he said.

To Kennedy’s credit, he’s never called french fries a health food. MAHA’s vision of an ideal food landscape is one where people eat more fresh fruits and vegetables, lean proteins, and minimally processed food (in addition to beef tallow and raw milk). Kennedy has long condemned processed foods and the companies that make them for poisoning Americans. Earlier this month, he lauded states for announcing plans to restrict SNAP recipients from using the benefits to buy candy and soda. Yet so far, his dealings with food companies themselves have been fundamentally friendly: asking them to voluntarily phase out food dyes, congratulating Coca-Cola for its commitment to sugar as a sweetener.

If Kennedy shies away from using the government’s real power, he could blow a genuine opportunity to overhaul America’s food landscape. Food companies have enormous power over what we eat and could effectively nudge Americans toward healthier habits, Bragg said. In the mid-aughts, for example, companies such as Campbell’s, Heinz, and Kraft committed to reducing salt levels in foods, including in breads, cold cuts, and cheese. It worked: From 2009 to 2018, the amount of salt in packaged food decreased by 8.5 percent. This outcome was partly driven by voluntary goals set by the National Sugar and Salt Reduction Initiative, a nongovernmental organization. The companies, however, also faced threats of regulation from the federal government if they did not comply. In 2016, the FDA proposed its own salt-reduction guidelines, further pressuring the food industry. “There has to be a threat of mandatory policy,” Moran said. “Otherwise, we’re just going to continue to see them making these changes around the margins that are very unlikely to meaningfully impact health.”

[Read: A ‘MAHA box’ might be coming to your doorstep]

Meanwhile, Kennedy’s HHS hasn’t instituted or threatened any binding regulations on food companies; indeed, it seems strongly opposed to doing so. A leaked draft of the second MAHA report, a document outlining HHS’s policy strategy that has yet to be finalized, explicitly details plans to deregulate food and agriculture. “The Trump administration has initiated a robust food policy agenda to Make America Healthy Again, from phasing out artificial food dyes to updating Dietary Guidelines for Americans to reforming the ‘Generally Recognized as Safe’ Standard,” the White House spokesperson Kush Desai told me in an email. (Under Kennedy, the FDA has so far revoked the authorization of one dye, Red 3. Formal changes to GRAS have not yet been announced.) “Every stakeholder in this movement—from parents to food companies to physicians to farmers to restaurants—has a role to play to transform how Americans view and make decisions about our health and nutrition.”

The superficial changes that companies have made to align with MAHA’s goals offer a glimpse of what could change if Kennedy were willing to enforce his more science-backed policy proposals. But as things stand, HHS is attempting to clean up America’s food supply with a spray bottle. What it really needs is a power washer.

From The Atlantic via this RSS feed

FBI directors don’t customarily announce raids in progress. But early this morning, Kash Patel celebrated the search of former National Security Adviser John Bolton’s home as agents were rolling into his suburban-Maryland driveway: “NO ONE is above the law … @FBI agents on mission,” Patel wrote on X. Agents also executed a search warrant at Bolton’s office in Washington, D.C. President Donald Trump later told reporters that he had learned about the raid on one of his most voluble critics from TV news, but he took the opportunity to call Bolton a “lowlife” and “not a smart guy.” Then he added: “Could be a very unpatriotic guy. We’re going to find out.”

The FBI’s actions were hard not to read as payback for Bolton’s years of criticism of the president, even as the facts that persuaded a judge to approve a search warrant remain unknown. That’s the problem with a politicized legal system—even if an investigation is legitimate, it’s easy to assume that its motives are corrupt. Trump has spent years vowing retribution against Bolton, particularly after Bolton published a 2020 memoir that portrayed the president as incompetent and out of his depth on foreign policy.

If this was revenge, it wasn’t an isolated act. As agents were still packing up boxes of Bolton’s effects, The Washington Post reported that Defense Secretary Pete Hegseth had pushed out yet another senior military officer, firing Lieutenant General Jeffrey Kruse, director of the Defense Intelligence Agency. In June, its analysts delivered a preliminary assessment that U.S. bombers had caused relatively limited damage to Iranian nuclear facilities, undercutting Trump’s pronouncements that the sites were “obliterated.” And just three days ago, Director of National Intelligence Tulsi Gabbard revoked the security clearances of more than three dozen current and former national-security officials. Several played key roles in efforts to counter or expose Russia’s 2016 election interference, what Trump calls the “Russia Hoax” and Gabbard has described as part of a “years-long coup” against the president.

[Read: The ‘Russia Hoax,’ Revisited]

Put it all together, and this may be remembered as the week Trump’s campaign against the “deep state” kicked into high gear. To some intelligence professionals I spoke with, it felt as though something fundamental had shifted in their historically apolitical line of work.

“Given the dystopian nature of it all—clearance revocations of former officials who did no wrong, forced retirements of long-standing intelligence officials, reductions in force that include junior officers who were just hired, and a wildly politicized leadership in the intelligence community—I no longer recommend young Americans to pursue careers in intelligence,” Marc Polymeropoulos, a veteran CIA officer who had his own security clearance yanked earlier this year, told me.

Purge doesn’t adequately capture what national-security experts see happening here. Chilling effect is too mild, though revoking the security clearances of two senior intelligence officers, as Gabbard did, effectively ending their government careers, will indeed send a message. Terrorizing the workforce is a phrase I heard a lot this week. And that may indeed be the point.

“Instead of being honest about what we think, now people will just keep their mouths shut or tell Trump what he wants to hear,” said one former official, who would only speak anonymously. The administration publicly identified this person as part of the “Russia Hoax,” and they’ve hired personal security for outside their home, fearing that Trump’s most fevered supporters might pay a visit.

Forget about calling out misbehavior or wrongdoing by administration officials, the person added: “Where would we go to file a grievance, or to report misconduct? Who’s going to do that?”

Gabbard’s office did not respond to my request for comment.

[Read: Tulsi Gabbard Chooses Loyalty to Trump]

One current official described the mood among career intelligence officers as “panicked.” In this person’s agency, three senior officials were abruptly placed on administrative leave this week. One of them has been involved in efforts to counter foreign threats against U.S. elections, which the administration has scaled back.

Gabbard’s actions have also raised concerns about separation of powers. She revoked the clearances of at least two congressional staffers. It will be difficult for them to perform their oversight of the executive branch without access to classified information.

Bolton was in his Washington office as the FBI conducted its search, according to a person close to him. He did not respond to a request for comment. Bolton was investigated during the first Trump administration and during the Biden administration over his book, The Room Where It Happened. He had submitted the manuscript for a prepublication review in early 2020, and after a lengthy back-and-forth with government officials, he made changes to address concerns about the possible disclosure of classified information. That effectively made it suitable for publication, according to a detailed statement from the official who led the review.

But in a highly unusual maneuver, the Trump White House ordered a second review by an administration official, who concluded that the manuscript was full of classified information. (That official, Michael Ellis, is now deputy director of the CIA.) The official in charge of the earlier review disagreed and concluded that the administration was trying to silence a political critic and was trampling his First Amendment rights.

Bolton published the book anyway. Federal investigators looked into whether he had illegally disclosed classified information. But Bolton was never charged. It’s possible some new evidence of a potential crime has emerged, leading to today’s FBI raid. But the administration’s hostility toward Bolton is well known, and Trump has made no secret of the fact that, seeing himself as the victim of political prosecutions during the Biden years, he is eager to turn the tables on perceived enemies. A senior U.S. official told the New York Post that the Biden administration had shut down the probe into Bolton “for political reasons.”

“That’s nonsense,” a former senior Justice Department official told me. “No decision in any case was ever made for political reasons. These accusations are obviously made in bad faith, and honestly, that’s what happens when you have people making decisions with basically no experience with complex national- security investigations. They have no clue what they’re talking about.”

There are still officials working in the government who took part in the 2016 efforts to counter Russia. Has the White House overlooked them? Are they next on the list to be purged? Everyone is left to wonder. But no one thinks that the president’s retribution campaign is anywhere near its end.

Vivian Salama and Isaac Stanley-Becker contributed reporting.

From The Atlantic via this RSS feed

This morning, Donald Trump claimed he knew nothing about the raid on the Maryland home of his former National Security Adviser John Bolton. “I’m not a fan of John Bolton,” Trump said, calling him a “lowlife.” In that respect he agrees with many Democrats: After Trump fired Bolton, in 2019, Representative Adam Schiff of California (now a senator) said that one “should question John Bolton’s patriotism,” and Representative Nancy Pelosi called him a “disgrace.” In January, Trump pulled Bolton’s security detail, which had been in place due to credible threats by the Iranian government, which is also not a fan of John Bolton. Bolton’s neighbor Gerald Rogell told The New York Times that Bolton was “unfriendly.” He added, dryly and a little gratuitously, “John Bolton is not our favorite person.” If a terrible accident befalls Bolton, one should not rule out a Murder on the Orient Express scenario. Who did it? Maybe everyone.

Reports say the government is searching Bolton’s home and office to determine whether he mishandled or shared classified material. (The New York Times says the raid followed “intelligence collected overseas” by the CIA.) The investigation is therefore unlikely to be related to Bolton’s long-standing dispute with Trump over Bolton’s memoir, which he published in 2020 without waiting for government clearance. (Bolton alleged that Trump’s administration slow-rolled the review to block criticism of the president.) Bolton has, ironically enough, been a longtime pooh-pooher of accusations that the federal government classifies way too much material, and that that secrecy-bloat impedes transparency.

[David A. Graham: John Bolton plumbs the depth of Trump’s depravity]

Bolton applies a lawyerly, punctilious zeal not only to his work but to many other aspects of his existence. Such people are not great at block parties. When you drop by to borrow sugar, they draft contracts with ruinous penalties and arbitration in the jurisdiction of Guantánamo Bay to ensure return of their measuring cup. In my 2019 profile of Bolton, I recounted his argument across the deli counter at a Safeway, over whether he was owed a refund for a turkey he had not yet consumed. Although these born sticklers may not be great neighbors, they tend to be scrupulous about personal liability, especially when they know—as Bolton must—that the president is gunning for them.

FBI agents were seen entering Bolton’s home with cardboard boxes. I wonder what they walked out with. Bolton is not the type to stumble into a situation that he cannot litigate his way out of. That said, he is cocky, and his memory is brimming with so many secrets that even a cautious person could spill one in an unguarded moment.

FBI Director Kash Patel tweeted, in apparent reference to the raid, that “NO ONE is above the law.” Patel has also recently pledged to “de-weaponize the FBI,” in particular by purging it and its partner agencies of those who have claimed that Trump has colluded with Russia. Bolton, in this regard, should be safe. He has used Trump’s own language in describing the Robert Mueller investigation as a “witch hunt,” and he has consistently denigrated Trump since 2020, not for treason but for ignorance, sloth, and foreign-policy blunders. After last week’s Ukraine summit between Trump and Russian President Vladimir Putin, Bolton accused Trump of one of the worst sins one can commit in Trump land: being low-energy. He told CNN that Trump “looked very tired up there,” and said one should “reflect on what that means.”

[Graeme Wood: John Bolton knows what he’s doing]

Reflection is certainly in order, on the topic of what it means that Patel is investigating a former Trump official just as that defector is getting more mordant in his criticisms of his former boss. The pledge to “de-weaponize” the FBI would be more credible from a less vindictive administration, and one that had not spent its opening months purging apolitical staff. Until more information comes out about the warrant and the intelligence behind it, no one can say definitively whether today’s raid is due to Bolton’s status as critic, Bolton’s bad judgment or malfeasance, or nothing at all. All remain possible. A truly depoliticized Justice Department and intelligence community—less politicized than Trump’s, and also less politicized than recent Democrats’—would leave less doubt about the relative likelihood of each. But that is not the Justice Department currently in place.

From The Atlantic via this RSS feed

This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.

Long flights can be tedious, but they also present a chance to watch movies both new and familiar. So we asked The Atlantic’s writers and editors: What is the perfect film to watch on a plane?

Crazy Rich Asians (available to rent on Prime Video and Apple TV+)

Sleep was elusive on my recent 15-hour flight. Never before had an airplane-movie formula been more important. First, the obvious: I look for anything that is mindlessly fun or otherwise engrossing enough to help me forget that I am hurtling through the air in a bus with wings. I probably will not watch anything worthy of the Criterion Collection. Provided that I am entertained, the movie’s length is no issue. And it doesn’t hurt if there are scenes involving airports or travel, to help me romanticize the experience of crying children and cramped seats.

These are among the many boxes that Crazy Rich Asians ticks. I’ve spaced out my rewatches of this movie enough that it still feels fresh and captivating every time. Michelle Yeoh’s performance as an icy, formidable matriarch is hard to look away from. The lavish scenery and shots of Singapore’s night-market food strengthen my travel urge. Plus, those sweet airplane scenes (no spoilers!). The only downside is that it makes me wish I no longer had to travel in economy.

— Stephanie Bai, associate editor

***

The Dark Knight Rises (streaming on HBO Max)

Moviegoing in an airplane may not elicit the same cinematic thrill as sitting in an IMAX theater, but it can bring a different emotional vulnerability. Perhaps you’ve just said goodbye to a lover you will never see again, or you’re flying home to fix a fraught situation with your parents. Feelings, too, can be heightened at cruising altitude.

I’ll just come out and say it: I’ve teared up more than once over the Atlantic Ocean while watching Bruce Wayne go to war with his inner demons in The Dark Knight Rises. The finale of Christopher Nolan’s morally serious and visually exquisite treatment of the Batman story culminates in Wayne’s showdown with a hulking Bane—ruthless and funnily accented, drunk off his own self-righteousness. I won’t give away the ending, but suffice to say, the moment when Alfred glances up from his caffè lungo in Florence is when I lose it.

— Thomas Chatterton Williams, staff writer

***

In Her Shoes (available to rent on Prime Video)

In Her Shoes, the 2005 dramedy, is white noise. No, pink noise—heartfelt but easy. Toni Collette and Cameron Diaz play feuding sisters. Shirley MacLaine is their long-lost grandmother. The movie has cute dogs and witty seniors. It’s an in-flight tonic that beats even benzos and cabernet.

For years, two colleagues and I would refer to In Her Shoes as a sacred text of procrastination; it was always playing on cable at just the right time. I asked them to endorse my nomination of the film.

“There are no banger lines you have to hear, no intricate plot points you have to absorb,” Monica told me. “It’s a low-altitude film to be enjoyed at high altitude.”

“It’s the best movie for just getting through something you don’t want to be doing,” Hank said. “Writing the rest of Chapter 7,” for example. Or: Sitting in seat 24C to LAX.

The film’s 20th anniversary is in early October. So, consider this blurb a commemoration. Cheers.

— Dan Zak, senior editor

***

The Fifth Element (available to rent on Prime Video and Apple TV+)

On a plane, I prefer to watch a movie that I already know well, and enjoy rewatching. Luc Besson’s The Fifth Element has always been a guilty pleasure, something I’ve seen so many times that I can fill in any dialogue, sound design, or music that might be missed because of noisy passengers or bad headphones.

The movie is unabashedly fun, and Bruce Willis has never been more charming than he is as a taxi driver tasked with saving the world from an evil sentient fireball. It’s easy to watch as an engaged viewer or as a distracted traveler, and the PG-13 rating makes it okay for public viewing. Sure, there is some silly dialogue (such as the phrase “slightly greasy solar atoms”) and subpar acting, but somehow that just adds to the appeal.

— Alan Taylor, senior photo editor

***

The movie the person near you is watching (playing on their screen)

The perfect airplane movie is the movie playing on one of your fellow passengers’ screens, which you watch through a crack between the seats or by glancing across the aisle. There’s an invigorating element of randomness (since you have no control over their selection), and also a pleasing surreptitious feeling about the whole experience, as if you’re getting away with something. Given that you won’t have any sound, subtitles are helpful, but sometimes it’s just as fun without them—consumed this way, even the most clichéd romantic comedy gains a certain art-house surrealism. Some things I have recently watched this way include Pulp Fiction, Dune: Part Two, multiple action films starring former professional wrestlers, and several episodes of the Batman miniseries The Penguin (out of order).

Later, on a different trip, I plugged in my headphones and watched Dune: Part Two with sound and dialogue on my very own screen. I have to say that it didn’t make much more sense.

— Quinta Jurecic, staff writer

***

Augmented-reality glasses (available to buy online)

Let’s be honest: The best airplane film is whatever film you personally want to watch. That’s probably going to be different for you than it is for me. But though I can’t tell you which movie to watch in midair, I can tell you how to watch the exact movie you want, regardless of what’s available on the seatback screen in front of you. All you need is a pair of augmented-reality glasses.

As somebody who travels a lot for work, I have perfected the art of bringing my own movie theater with me. Unlike their more famous and expensive VR headset counterparts, these types of glasses are lightweight and won’t leave you exhausted from wearing them or looking like an alien to passersby. Most important, the glasses can connect to your smartphone, laptop, or video-game device, meaning that if you can stream it or download it, you can project it onto your own giant, private silver screen. Companies such as Xreal, Viture, and Rokid offer a range of models, features, and prices. Some even allow you to adjust the display for your prescription. Find the right one for you—you can typically get the same functionality for less money if you buy an old model—and you’ll never fly the same way again.

— Yair Rosenberg, staff writer

Here are three new stories from The Atlantic:

MAGA world is so close to getting it, Tom Nichols writes.The states where it’s riskier to have a babyThe West Bank is sliding toward a crisis.

Today’s News

The FBI searched the home and office of John Bolton, a former national security adviser to President Donald Trump, as part of an investigation into whether he illegally shared or possessed classified information, according to sources familiar with the investigation.A federal judge ordered Florida last night to halt construction on the “Alligator Alcatraz” immigrant-detention facility in the Everglades and to stop bringing in new detainees. The judge also ruled that within 60 days, “all generators, gas, sewage, and other waste and waste receptacles” must be removed, citing environmental damage.Canada announced that it will remove its 25 percent retaliatory tariffs from about half of the U.S. goods it has targeted this year, but Canadian Prime Minister Mark Carney confirmed that duties on U.S. steel, aluminum, and automobiles will remain.

Dispatches

The Books Briefing: For some, including the author Lauren Groff, travel remains a spiritual endeavor, Boris Kachka writes. Read more about her search for an 11th-century novelist in Kyoto.The Weekly Planet: The model that once saved the New Jersey Meadowlands is disappearing, and legal protections for such places are now under threat, Kyra Morris writes.

Explore all of our newsletters here.

Evening Read

Collection of the Smithsonian National Museum of African American History and Culture

Actually, Slavery Was Very Bad

By Clint Smith

In what looks to be an intensifying quest to reshape American history and scholarship according to his own preferences, President Donald Trump this week targeted the Smithsonian Institution, the national repository of American history and memory. Trump seemed outraged, in particular, by the Smithsonian’s portrayal of the Black experience in America. He took to Truth Social to complain that the country’s museums “are, essentially, the last remaining segment of ‘WOKE.’ The Smithsonian,” he wrote, “is OUT OF CONTROL.” Then Trump wrote something astonishing, even for him. He asserted that the narrative presented by the Smithsonian is overly focused on “how horrible our Country is, how bad Slavery was, and how unaccomplished the downtrodden have been.”

Before continuing, it is important to pause a moment and state this directly: Donald Trump, the current president of the United States, believes that the Smithsonian is failing to do its job, because it spends too much time portraying slavery as “bad.”

Read the full article.

More From The Atlantic

Autocracy in America: The view from NATO’s eastern flankThe culture war over nothingWhat Trump doesn’t understand about “America First”

Culture Break

Photo-Illustration by Ben Kothe / The Atlantic. Source: Hulu / Everett Collection.

Watch. The sensitive tween of the animated show King of the Hill is now an adult—with extremely Millennial anxieties, Jeremy Gordon writes.

Take a look. These photos of the week show a sail-in parade in Amsterdam, harness racing in Germany, an independence celebration in Indonesia, and more.

Play our daily crossword.

Rafaela Jinich contributed to this newsletter.

When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic*.*

From The Atlantic via this RSS feed

Something strange has been happening on YouTube over the past few weeks. After being uploaded, some videos have been subtly augmented, their appearance changing without their creators doing anything. Viewers have noticed “extra punchy shadows,” “weirdly sharp edges,” and a smoothed-out look to footage that makes it look “like plastic.” Many people have come to the same conclusion: YouTube is using AI to tweak videos on its platform, without creators’ knowledge.

A multimedia artist going by the name Mr. Bravo, whose YouTube videos feature “an authentic 80s aesthetic” achieved by running his videos through a VCR, wrote on Reddit that his videos look “completely different to what was originally uploaded.” “A big part of the videos charm is the VHS look and the grainy, washed out video quality,” he wrote. YouTube’s filter obscured this labor-intensive quality: “It is ridiculous that YouTube can add features like this that completely change the content,” he wrote. Another YouTuber, Rhett Shull, posted a video last week about what was happening to his video shorts, and those of his friend Rick Beato. Both run wildly popular music channels, with more than 700,000 and 5 million subscribers, respectively. In his video, Shull says he believes that “AI upscaling” is being used—a process that increases an image’s resolution and detail—and is concerned about what it could signal to his audience. “I think it’s gonna lead people to think that I am using AI to create my videos. Or that it’s been deepfaked. Or that I’m cutting corners somehow,” he said. “It will inevitably erode viewers’ trust in my content.”

Fakery is a widespread concern in the AI era, when media can be generated, enhanced, or modified with little effort. The same pixel-filled rectangle could contain the work of someone who spent time and energy and had the courage to perform publicly, or of someone who sits in bed typing prompts and splicing clips in order to make a few bucks. Viewers who don’t want to be fooled by the latter must now be alert to the subtlest signs of AI modification. For creators who want to differentiate themselves from the new synthetic content, YouTube seems interested in making the job harder.

[Read: ChatGPT turned into a Studio Ghibli machine. How is that legal?]

When I asked Google, YouTube’s parent company, about what’s happening to these videos, the spokesperson Allison Toh wrote, “We’re running an experiment on select YouTube Shorts that uses image enhancement technology to sharpen content. These enhancements are not done with generative AI.” But this is a tricky statement: “Generative AI” has no strict technical definition, and “image enhancement technology” could be anything. I asked for more detail about which technologies are being employed, and to what end. Toh said YouTube is “using traditional machine learning to unblur, denoise, and improve clarity in videos,” she told me. (It’s unknown whether the modified videos are being shown to all users or just some; tech companies will sometimes run limited tests of new features.)

Toh’s description sounds remarkably similar to the process undertaken when generative-AI programs create entirely new videos. These programs typically use a diffusion model: a machine-learning program that is trained to refine an extremely noisy image into one that’s clear, with sharp edges and smooth textures. An AI upscaler can use the same diffusion process to “improve” an existing image, rather than to create a new one. The similarity of the underlying process might explain why the visual signature of diffusion-based AI is recognizable in these YouTubers’ videos.

While running this experiment, YouTube has also been encouraging people to create and post AI-generated short videos using a recently launched suite of tools that allow users to animate still photos and add effects “like swimming underwater, twinning with a lookalike sibling, and more.” YouTube didn’t tell me what motivated its experiment, but some people suspect that it has to do with creating a more uniform aesthetic across the platform. As one YouTube commenter wrote: “They’re training us, the audience, to get used to the AI look and eventually view it as normal.”

Google isn’t the only company rushing to mix AI-generated content into its platforms. Meta encourages users to create and publish their own AI chatbots on Facebook and Instagram using the company’s “AI Studio” tool. Last December, Meta’s vice president of product for generative AI told the Financial Times that “we expect these AIs to actually, over time, exist on our platforms, kind of in the same way that [human] accounts do.”

[Read: What we discovered on “deep YouTube”]

In a slightly less creepy vein, Snapchat provides tools for users “to generate novel images” of themselves based on selfies they’ve taken. And last year, TikTok introduced Symphony Creative Studio, which generates videos and includes a “Your Daily Video Generations” feature that suggests new videos automatically each day.

This is an odd turn for “social” media to take. Platforms that are supposedly based on the idea of connecting people with one another, or at least sharing experiences and performances—YouTube’s slogan until 2013 was “Broadcast Yourself”—now seem focused on getting us to consume impersonal, algorithmic gruel. Shull said that the modification of his videos erodes his trust in YouTube, and how could it not? The platform’s priorities have clearly shifted away from creators such as Shull, whose combined work is a major reason YouTube has become the juggernaut it is today.

From The Atlantic via this RSS feed

This is an edition of the Books Briefing, our editors’ weekly guide to the best in books. Sign up for it here.

One of the more common clichés of modern travel is calling any trip—even a subway ride to an Instagram-famous coffee shop—a pilgrimage. The word originally applied to journeys made to holy places by people so devout that they were willing to endanger their lives to get there. Today, both the risks and rewards of travel tend to be lower, but the activity retains its spiritual character for some, including the novelist Lauren Groff. For the latest installment of The Atlantic’s series “The Writer’s Way,” she traveled to Kyoto in search of the mysterious author of The Tale of Genji, frequently credited as the world’s first novel. She made her way through the crowds swarming Japan’s former imperial capital to find out more about that writer, known to us as Lady Murasaki. But Groff also came across the kinds of spiritual experiences that fire up much of her own fiction.

First, here are four new stories from The Atlantic’s books section:

She has no autonomy. Can she be happy?These books won’t make you a better person.“Paragraph,” a poem by Richard SikenSix books to read before you get to the airport

Groff was in Kyoto in April; the journalist Reeves Wiedeman was there around the same time. In a feature published in June in New York magazine, Wiedeman wrote that the city has become the epicenter of the “age of overtourism”: a once-tranquil historical landmark blighted by travelers racing to take selfies at a handful of clogged sites. Reading it, I wondered how Groff’s essay could wrest meaning from this location—what weaving among the frequent-flying box-checkers could reveal about the Heian era of Japan, a time and place that Groff says is “thrillingly distant to my imagination.”

In Kyoto, Groff did what many tourists do: She made a list and checked off destinations—temples, palaces, and museums associated with Murasaki’s life and work. Yet her most meaningful encounters had as much to do with sensation as place. She describes a feeling of “living outside time” while eating a 7-Eleven egg sandwich and sitting on a clean-swept sidewalk curb; she has an epiphany not while beholding a 10th-century relic but while taking a hot bath downstairs from her hotel room. Her deepest connections to medieval Japan are experiential, rather than physical or intellectual. “I had an inkling that, though my love of Lady Murasaki could be explained only through beautiful abstraction—by meeting her mind in her work,” Groff writes, “I might begin to understand something tangible about her through the wordless animal body.”

This kind of sensory awareness can be found in Groff’s fiction. Her most explicitly religious novel, Matrix, published in 2021, imagined the 12th-century mystic Marie de France as a towering figure who made a British abbey into a power center for medieval women. A heterodox interpretation of Christianity infuses much of her work, as Judith Shulevitz noted in a recent Atlantic essay about her latest novel, The Vaster Wilds. Shulevitz considered the journey in the book, a young woman’s flight from Jamestown in the 17th century, to be a spiritual one—an update, in fact, of The Pilgrim’s Progress—in which communion with nature is achieved through perilous struggle. She called the book “Christian allegory in a post-Christian spirit.” Groff’s recent novels, as my colleague Sophie Gilbert wrote in a profile when Matrix was released, sprang from “the idea that so much of our present suffering comes from a misreading of Genesis. God instructed man to have dominion over Earth and its creatures, and yet dominion, Groff thinks, has been interpreted as domination instead of care: ‘the right to kill, the right to take, and not the right to nurture.’”

I don’t think it’s a stretch to connect this dichotomy—dominion versus care—with the approach Groff takes in Kyoto, diverging from the flocking tourists that Wiedeman depicts. Groff, a fan (as Shulevitz notes) of the animist-leaning Quaker John Bartram, observes the nature-worship of Japan’s Shinto traditions. She closes her essay with a tea-and-meditation ceremony at the Shunkō-in Temple, a place with no ostensible connection to Murasaki, and yet she gleans something valuable about the often-puzzling structure of The Tale of Genji. She learns from one of the temple’s Buddhist reverends that “the self is a shifting, inconstant phenomenon”; he advises her to “embrace” ambiguity, which is “part of nature.” This instruction helps Groff understand the orderly disorder of Murasaki’s writing; it also teaches her about herself. Perhaps this is—or should be—the goal of every pilgrimage.

Takako Kido for The Atlantic

A Tale of Sex and Intrigue in Imperial Kyoto

By Lauren Groff

A thousand years ago, Murasaki Shikibu wrote The Tale of Genji, the world’s first novel. Who was she?

Read the full article.

What to Read

The Deltoid Pumpkin Seed, by John McPhee

Pilots get most of the public credit for a flight’s successes—but they couldn’t go anywhere without the behind-the-scenes heroes: engineers. McPhee has a rare gift for stepping into the astonishing obsessions of seemingly ordinary working people; here, he uses it to immerse the reader in a decades-long quest to build an entirely new type of aircraft. That potential vehicle, shaped like the titular pumpkin seed, was imagined as a combination of dirigible and airplane. Its siren call, as McPhee shows, was sometimes all-consuming, even life-destroying. In a saga that reaches from the Civil War to the 1970s, one acolyte after another grew convinced that he (this affliction appears to target men exclusively) would be the one who conquered the engineering challenge that had theretofore led only to ruin. Did anyone finally succeed? The fact that you aren’t reading these words in the passenger compartment of a dirigible-airplane hybrid gives you a clue, but McPhee’s storytelling makes readers hope that the mission will somehow pan out.  — Jeff Wise

From our list: Six books to read before you get to the airport

Out Next Week

📚 A New New Me, by Helen Oyeyemi

📚 The Second Emancipation: Nkrumah, Pan-Africanism, and Global Blackness at High Tide, by Howard W. French

📚 Breakneck: China’s Quest to Engineer the Future, by Dan Wang

Your Weekend Read

Illustration by Jonelle Afurong / The Atlantic. Sources: Mike Hansen / Getty; mikroman6 / Getty.

Why Is Everything Spicy Now?

By Ellen Cushing

To put it generally and reductively, American food has not always been known for embracing spice. But now a large and apparently growing number of people in this country are willingly chomping down on fruits that have been expressly cultivated to bind to their body’s pain receptors and unleash fury with every bite. “It’s one of the great puzzles of culinary history,” Paul Rozin, a retired psychologist who spent much of his career studying spice, told me. “It is remarkable that something that tastes so bad is so popular.”

Read the full article.

When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic*.*

Sign up for The Wonder Reader, a Saturday newsletter in which our editors recommend stories to spark your curiosity and fill you with delight.

Explore all of our newsletters.

From The Atlantic via this RSS feed