Photographs by Jenn Ackerman

When Melissa Kono, the town clerk in Burnside, Wisconsin, began training election workers in 2015, their questions were relatively mundane. They asked about election rules, voter eligibility, and other basic procedures. The job was gratifying and enjoyable; they helped their neighbors while sipping coffee.

But over the past few years, everything has changed. Kono now finds herself fielding questions about what to do when approached by suspicious voters who ask provocative questions or gripe about fraud. She’s added an entire training section dedicated to identifying threats and how to report them. “I never in a million years imagined that that would be part of my curriculum,” she told me. Kono has yet to receive any direct threats herself—perhaps, she thinks, because Donald Trump won the popular vote in her area in 2016 and 2020—but she fears that things may be different this time around. “What I do hear is I know the election is not rigged here, but in other places,” she said. “And I’m honestly worried sometimes: What if Harris wins? What if it gets too close? And now they start questioning me or coming after me, when I have nothing to do with the outcome.”

Around the country, election officials have already received death threats and packages filled with white powder. Their dogs have been poisoned, their homes swatted, their family members targeted. In Texas, one man called for a “a mass shooting of poll workers and election officials” in precincts with results he found suspicious. “The point is coercion; the point is intimidation. It’s to get you to do or not do something,” Al Schmidt, the secretary of the Commonwealth of Pennsylvania, told me—to get you to “stop counting votes, or we’re going to murder your children, and they name your children,” a threat that Schmidt said he received in 2020. This year, the same things may well happen again. “I had one election official who said they called her on her cellphone and said, ‘Looks like your mom made lasagna tonight; she’s wearing that pretty yellow dress that she likes to wear to church,” Tammy Patrick, the chief programs officer at the National Association of Election Officials and a former elections officer in Maricopa County, Arizona, told me. “It’s terrorism here in America.”

These workers, from secretaries of state to local officials to volunteers, are bearing the immediate, human toll of a campaign to discredit the integrity of American democracy. They are the most direct and vulnerable targets for people who have embraced conspiracy theories about fraudulent and “stolen” votes following the 2020 election—unfounded claims that have been directly promoted by Trump and many other members of the Republican Party, who still will not accept that he lost his first reelection bid. Where candidates used to compete against each other, Schmidt told me, some are now “attacking the referees.” In the most extreme narratives, election workers are accused of fabricating, shredding, or double-counting ballots, which leads to suspicion and harassment. “Since the 2020 election, we have seen an unprecedented spike in threats against the public servants who do administer our elections,” including shootings and a bomb threat, Attorney General Merrick Garland said last month. A survey conducted in February and March by the Brennan Center for Justice found that 38 percent of election officials reported being harassed, abused, or threatened—up from 30 percent a year earlier.

Election workers take notes during training at the American Legion in Neillsville, Wisconsin, on Thursday, October 24, 2024. (Jenn Ackerman for The Atlantic)

This is not just a story about assaults on individual workers, although that would be bad enough. Election administration is “underappreciated as the foundation upon which all of our representative government thrives,” Rachel Orey, the director of the Bipartisan Policy Center Elections Project, told me. In a very real way, these officials represent the soul of democracy. Many of them are undertaking their duties while also juggling child care and everyday errands such as grocery shopping. Without their diligence, nobody could be elected, period. The form of government that Americans recognize and celebrate could not exist.

Dissuading or preventing people from going to, or otherwise attempting to interfere with, the polls are century-old dirty tactics, and there are all manner of legal ways to suppress or dilute the vote, many of which target racial minorities. But Trump’s attempts to unilaterally dictate election results are different. As far back as 2012, he criticized Barack Obama’s reelection as a “total sham and a travesty.” Victory in 2016, and the conversion or defeat of nearly all of his Republican rivals, gave Trump the power to mount a serious and systematic attempt to discredit the democratic process. He and his furious supporters, in turn, have unleashed a sustained assault on national and state elections alike.

“I wasn’t aware of any real threats or harassment or wide-scale verbal abuse on election officials prior to the 2020 election cycle,” Tina Barton, the vice chair of the Committee for Safe and Secure Elections, told me. The outrage and attacks that emerged in 2020 have now been harnessed into a well-funded campaign: Republicans have reportedly donated upwards of $100 million to a network of so-called election-integrity groups to lay the groundwork for contesting the results should Trump lose again. Although poll watching is itself normal, the GOP is training and deploying armies of monitors with the presumption of fraud, flooding election offices with public-records requests, and filing endless challenges to voter-registration records. “I don’t think there’s any question that there is a more coordinated and sophisticated effort ahead of this election to discredit it than there was in 2020,” Lawrence Norden, the vice president of the elections and government program at the Brennan Center, told me.

Jenn Ackerman for The Atlantic

This election cycle, one of the “dominant narratives” is about noncitizen voting, Thessalia Merivaki, a political scientist at Georgetown University who studies how election officials combat misinformation, told me. Republican activists, politicians, lawmakers, and pundits have especially seized on false fears about immigrant and foreign voting to burnish a conspiracy theory that noncitizen votes from overseas will turn the election. These claims have been widely debunked, but all the energy behind them could delay vote counts and disenfranchise citizens. Republican groups are filing more and more lawsuits in battleground states about voter-identification requirements, absentee ballots, and other basic procedures, “setting up an opportunity afterwards to cast doubt on the election results,” Norden said.

When I began reporting this article, I was curious about whether election officials had concerns over new technologies. The internet has changed substantially since 2020. For the past couple of years, I’ve written about the rise of generative AI and its attendant issues: In terms that are most directly relevant to the election, that means the arrival of easy-to-create and highly convincing deepfakes, the concoction of micro-targeted conspiracy theories, the overall degradation of our information environment, and the possibility that our sense of shared reality might be wiped out altogether. It’s easy to imagine that AI could wreak havoc around Election Day—earlier this year, a robocall that cloned President Joe Biden’s voice was used in a voter-suppression effort—and experts have made their concerns about the technology clear.

The election workers and officials I spoke with did express worry about AI and its ability to accelerate disinformation and election-interference campaigns. But they also described problems that came from more familiar sources. They spoke with me about how videos of entirely proper and legal election procedures—snippets of livestreamed election procedures, for instance—had been miscontextualized to suggest that officials who were simply following the rules were actually smuggling in ballots, rigging voting machines, or otherwise manipulating the results. Blatantly false headlines and incendiary posts spreading on messaging apps and among social-media groups have done and continue to do plenty of damage. Sometimes, the details are irrelevant. As my colleague Charlie Warzel recently wrote, manipulated media and misinformation is useful not necessarily because it convinces some population of undecided suckers, but because it allows the already aggrieved to sequester themselves in a parallel reality. A voter might say, “Let’s just set the facts aside,” Patrick, of the National Association of Election Officials, told me, “and I’m going to tell you what I think or what I feel about this.”

Amy Burgans, the clerk-treasurer in Douglas County, Nevada, herself had doubts about the outcome of the 2020 election when she started in her role that December. (The previous clerk-treasurer had resigned—the pandemic and contentious election cycle, Burgans told me, had been “a lot.”) Burgans said that she had heard in the news, on social media, and from people she knew that there “must have been” some foul play. But once she was in charge of running elections and administered the 2022 midterms, she said, she saw the rigor at every step of the process and understood that the allegations of widespread, systemic fraud were impossible.

Now she’s the one fielding questions, frequently about voting machines. Burgans has explained to voters all of the controls in place, that she’s “never seen even one error” after an election audit. Still, people ask her about election procedures at almost every event she attends, or “even if I’m at the Elks Club just hanging out,” she said. Burgans said she has “no issues with” and tries to address these questions. But doing so takes time and energy—and these comments are just the tip of the spear.

Burgans received a threat in the mail in 2022. Although it was mostly a broad rant about the government, it did make her worry for her children’s safety, and she installed a security system in her home. Her county’s election facilities are stocked with personal protective equipment and Narcan, in the event of suspicious substances or powders (which might be Fentanyl) arriving in the mail—something that has already happened at election offices in several states. She also recently installed bulletproof glass in the office, where Burgans and full-time staff work—as a precaution rather than a response to any particular threat, she said. Election deniers are “consistently coming into [election] offices saying things like ‘You’d better watch your back’ or ‘Don’t you forget: I know where your kids go to school,” Barton, the Committee for Safe and Secure Elections vice chair, told me. What was unprecedented in 2020, Barton said, is now an “ongoing onslaught.”

These workers, from secretaries of state to local officials to volunteers, are bearing the immediate, human toll of a campaign to discredit the integrity of American democracy. (Jenn Ackerman for The Atlantic)

This attack on American elections is not an invasion so much as a siege. And just as hateful, outlandish, and conspiracist misinformation have eroded Americans’ trust in one another, institutions, and basic facts, this environment is taking a psychic toll on election workers. They find themselves having to put in extra hours to field questions, accommodate an influx of poll watchers, process voter challenges, sort through public-records requests, and prepare for any emergencies and attacks—all while fearing for their safety. For more than two decades, running an election has become steadily more complex and involved. After 2000’s infamous hanging chads, election workers had to become IT professionals. After long lines became a key issue in 2008 and 2012, they became logistics experts. After 2016, they learned cybersecurity, and in advance of 2020, they studied public-health protocols and how to process enormous quantities of mail-in ballots. Now election workers have to be communications experts as well. “We’ve had poll watchers in here every single day since September 26,” when early voting began, “sometimes three or four of them in a small space,” Aaron Ammons, the clerk and recorder of deeds in Champaign County, Illinois, said in a recent press briefing.

Meanwhile, support and resources for these emerging responsibilities are frequently missing. The result, inevitably, is burnout: The job keeps getting harder and requiring more hours, but resources for hiring, buying new equipment, improving security, and more have been inconsistent and haphazard. “There have been new challenges and new expectations put on election administrators, but funding hasn’t kept pace,” Rachel Orey said. Hours spent on election work have ballooned since 2020, according to a recent national survey of election workers conducted by Reed College. Meanwhile, nearly one-third of election offices don’t have any full-time staff, wages are pitiful, and turnover rates grew from 28 percent in 2004—already high—to nearly 39 percent in 2022.

This burden “has taken away from [election officials’] ability to just focus on the mechanics of that very important election,” Kim Wyman, a former secretary of state for Washington who recently served as a senior election-security adviser at the Department of Homeland Security, told me. Skeptics will use innocent mistakes and logistical snares—which are mundane and easily rectified—and even the act of correcting “as gasoline on the fire to ‘prove’ their point or their claim of voter fraud,” she said. This, in turn, only fuels the exhaustion. “People just make up stuff about what we do and are coming after us,” Kono, of Wisconsin, told me. She’s seen many longtime clerks and election workers leave, telling her, “I can’t do another presidential election” and “I don’t want to have to deal with voters.”

Those who remain do not take the job lightly. In 20 years working in election administration, Barton told me, “I have never seen election officials train so much in four years’ time”—improving security, being transparent at every step of the process, speaking at events and posting on social media to educate their communities. They’ve been preparing for November 5, 2024, for four years, Wyman told me. “This is my Olympics,” Kono said.

The misinformation crisis is commonly understood as a clash between two “realities” that is most visible online, in the words of high-profile politicians, or during spectacular flashpoints such as the January 6 Capitol riot. But for four years, and especially in the weeks leading up to and after November 5, these battles have and will be quotidian and interpersonal. “We are your soccer coaches. We are the moms helping at the schools, the dads coaching baseball, the grandmothers that are going on field trips,” Burgans said. This everyday warfare, waged against the neighbors and teachers and elders and bus drivers who administer the polls, and in turn democracy, may be more consequential than any single vote or outcome.

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.

Over the past three months, Elon Musk has mobilized his many resources—his exceptional wealth, far-reaching online platform, and time—for a cause that could have profound effects on his personal fortune and American society: electing Donald Trump.

Musk is going all in: In addition to donating $75 million to America PAC, a group he founded that backs Trump, he has also temporarily relocated to the all-important swing state of Pennsylvania to effectively run Trump’s get-out-the-vote strategy from a war room he set up in Pittsburgh. He has stumped on the trail, hosting a Trump town hall in the auditorium of a Pennsylvania high school last week and telling locals to go “hog wild” on voter registration. And, in his latest stunt, he has offered $1 million a day to registered voters in swing states who sign an America PAC petition backing the First and Second Amendments—a move that the Justice Department reportedly said might be breaking election laws. His efforts may prove consequential: As my colleague Franklin Foer wrote this past weekend, “If Trump wins, it will likely be by a narrow margin that can be attributed to turnout. Musk can tout himself as the single variable of success.”

Musk is far from the only major donor in this race. Bill Gates has reportedly given $50 million to Vice President Kamala Harris’s campaign, and various billionaires publicly support Harris or Trump. What distinguishes Musk though, beyond his on-the-ground efforts, is his ownership of X. He can spread information (and disinformation) with ease, and stifle views he doesn’t like, Sophia Rosenfeld, a historian at the University of Pennsylvania, told me in an email. Media owners have always been influential in American politics (Rupert Murdoch, for example, played a prominent role in past elections through his leadership of Fox News). But Rosenfeld noted that Musk’s particular combination of wealth and media control is “unprecedented.”

Musk’s audience is massive on X: His posts, many of which have amplified false and inflammatory rhetoric, get billions of views. Over the weekend he boosted the baseless claim that Michigan had more registered voters than eligible citizens. After Michigan Secretary of State Jocelyn Benson said that wasn’t true—and that Musk was spreading “dangerous disinformation”—Musk doubled down and accused her of lying to the public. This disinformation had a swift real-world impact: Benson told CBS that her team received harassing messages and threats after Musk’s post. Such rhetoric has the potential to warp how much voters trust election processes. Musk’s America PAC has also been urging people to report examples of “voter fraud” through what it calls the Election Integrity Community on X. Though such fraud remains exceptionally rare, his efforts could further sow distrust in election integrity and lay the groundwork for future claims of a stolen race. (America PAC did not immediately respond to my request for comment.) So prominent is Musk’s role in the MAGA movement that Minnesota Governor Tim Walz joked archly at a recent rally: “I’m going to talk about [Trump’s] running mate …. Elon Musk.”

Musk wasn’t always aligned, at least in public, with such zealotry. He reportedly said that Trump was a “stone-cold loser” in 2020, and he supported Barack Obama in 2008 and 2012. Still, as my colleague Charlie Warzel told me last month, Musk’s feelings of being aggrieved and attacked escalated when he faced pushback from liberals after his Twitter takeover; soon after, he began using X as a megaphone for MAGA. And, though his Trump endorsement seemed out of step with his long-standing image as a climate innovator, it is consistent with his rightward drift: Over the past few years, he has reportedly been quietly donating to Republican causes and candidates, including giving $10 million to Florida Governor Ron DeSantis last year for his ill-fated primary run.

The wealthy have long played an outsize role in politics—but Musk, as he so often does, is venturing to new extremes. If Trump wins, Musk’s gamble may pay off handsomely: In addition to a promised role in Trump’s government, he is poised to receive epic government contracts for his companies. But even if Trump doesn’t win, Musk could set a precedent for uber-rich donors getting more directly involved with political campaigns; that could intensify the “oligarchic side of modern American democracy,” Rosenfeld warned. Though Musk’s hands-on, incendiary campaigning methods are chaotic—and possibly illegal—his efforts during this election may pioneer a model for other megadonors looking to reshape a race.

Related:

What Elon Musk really wants Elon Musk has reached a new low.

Here are three new stories from The Atlantic:

If Trump loses, will his supporters believe it? Are you a Platonist or an Aristotelian? Michel Houellebecq has some fresh predictions. Be afraid.

Today’s News

Secretary of State Antony Blinken said that U.S. and Israeli negotiators will travel to Qatar in the coming days for Gaza cease-fire talks. Former President Barack Obama joined Kamala Harris at a rally in Atlanta tonight. A Los Angeles prosecutor is recommending the resentencing of Erik and Lyle Menendez, who were convicted in 1996 for the murder of their parents, after new evidence surfaced suggesting that their father sexually abused them.

Dispatches

The Weekly Planet: Cheap solar panels are changing the world, Zoë Schlanger writes. Time-Travel Thursdays: The myths that fueled marijuana’s criminalization have deep roots, Malcolm Ferguson writes.

Explore all of our newsletters here.

Evening Read

Illustration by Paul Spella / The Atlantic. Source: Getty.

Ratpocalypse Now

By Annie Lowrey

Has any man in history talked about “how much he hates rats” more than New York City Mayor Eric Adams? Adams himself posed that question at the city’s inaugural National Urban Rat Summit last month. “Let’s figure out how we unify against public enemy number one: Mickey and his crew.”

Mickey is, canonically, a mouse. But Adams’s campaign against the city’s endemic brown-rat population might be the most effective and highest-profile initiative of his scandal-ridden mayoralty.

Read the full article.

More From The Atlantic

The swing states are in good hands, Paul Rosenzweig argues. The dilemma at the heart of McDonald’s E. Coli outbreak America’s strangest tourist destination

Culture Break

AA Film / Moviestore / Columbia / AF Archive / Allstar / Alamy; Adam Maida / The Atlantic

Watch. Settle in with one of these 26 movies that critics were wrong about.

Read. These seven true stories read like thrillers.

Play our daily crossword.

Stephanie Bai 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.

This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.

The list of Nobel laureates reads like a collection of humanity’s greatest treasures: Albert Einstein, Marie Curie, Francis Crick, Toni Morrison. As of this morning, it also includes two physicists whose research, in the 1980s, laid the foundations for modern artificial intelligence.

Earlier today, the 2024 Nobel Prize in Physics was awarded to John Hopfield and Geoffrey Hinton for using “tools from physics to develop methods that are the foundation of today’s powerful machine learning.” Hinton is sometimes referred to as a “godfather of AI,” and today’s prize—one that is intended for those whose work has conferred “the greatest benefit to humankind”—would seem to mark the generative-AI revolution, and tech executives’ grand pronouncements about the prosperity that ChatGPT and its brethren are bringing, as a fait accompli.

Not so fast. Committee members announcing the prize, while gesturing to generative AI, did not mention ChatGPT. Instead, their focus was on the grounded ways in which Hopfield and Hinton’s research, which enabled the statistical analysis of enormous datasets, has transformed physics, chemistry, biology, and more. As I wrote in an article today, the award “should not be taken as a prediction of a science-fictional utopia or dystopia to come so much as a recognition of all the ways that AI has already changed the world.”

AI models will continue to change the world, but AI’s proven applications should not be confused with Big Tech’s prophecies. Machines that can “learn” from large datasets are the stuff of yesterday’s news, and superintelligent machines that replace humans remain the stuff of yesterday’s novels. Let’s not forget that.

Illustration by The Atlantic. Source: Science & Society Picture Library / Getty.

AI’s Penicillin and X-Ray Moment

By Matteo Wong

Today, John Hopfield and Geoffrey Hinton received the Nobel Prize in Physics for groundbreaking statistical methods that have advanced physics, chemistry, biology, and more. In the announcement, Ellen Moons, the chair of the Nobel Committee for Physics and a physicist at Karlstad University, celebrated the two laureates’ work, which used “fundamental concepts from statistical physics to design artificial neural networks” that can “find patterns in large data sets.” She mentioned applications of their research in astrophysics and medical diagnosis, as well as in daily technologies such as facial recognition and language translation. She even alluded to the changes and challenges that AI may bring in the future. But she did not mention ChatGPT, widespread automation and the resulting global economic upheaval or prosperity, or the possibility of eliminating all disease with AI, as tech executives are wont to do.

Read the full article.

What to Read Next

Today’s Nobel Prize announcement focused largely on the use of AI for scientific research. In an article last year, I reported on how machine learning is making science faster and less human, in turn “challenging the very nature of discovery.” Whether the future will be awash with superintelligent chatbots, however, is far from certain. In July, my colleague Charlie Warzel spoke with Sam Altman and Ariana Huffington about an AI-based health-care venture they recently launched, and came away with the impression that AI is becoming an “industry powered by blind faith.”

P.S.

A couple weeks ago, I had the pleasure of speaking with Terence Tao, perhaps the world’s greatest living mathematician, about his perceptions of today’s generative AI and his vision for an entirely new, “industrial-scale” mathematics that AI could one day enable. I found our conversation fascinating, and hope you will as well.

— Matteo

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.

In our scattered social-media age, a strange little problem has emerged: It is hard to figure out how to invite people to a party. A slew of digital tools is available—Paperless Post, Instagram stories, Partiful, a simple group text—that should theoretically make it easier to reach people. But it also means you have no one way to contact everyone you want to invite, and you’re left sending out multiple emails and posts for a single party. Sending a few extra texts hardly ranks among the world’s most pressing problems, but finding ways to gather people together is a meaningful act during a time when so many Americans—even the ones with friends—deal with loneliness and isolation. The challenge of the modern party invite is the story of the changing web in miniature: In recent decades, everyone seemed to be hanging out in the same few places online. Now people are dispersed widely across platforms, with even more variability based on age and affinity.

If you are loyal to a particular invitation method—or if you simply call your friends when you’re having a party—you may be scratching your head, wondering what I am talking about here. That’s fair enough. But according to my unscientific surveys, I am not the only one living in a dispersed invitation landscape. The other day, I texted a group of family members asking in what formats they get invited to parties. My Boomer mom responded first: Paperless Post, always, she said. My Gen Z sister, scoffing at the idea of receiving an email invite, said she mostly gets invites via the app Partiful, or group texts with friends (“grexts,” as she called them). My Millennial sister-in-law said she usually receives emails or texts from friends. Another, just as unscientific, poll of my colleagues indicated a similarly diverse range of invite approaches.

Geography seems to play a part too: My East Coast colleagues, especially those based in New York City, were familiar with Partiful, whereas that name meant little to people in other regions. (Partiful declined to share information about its users’ age and regional distribution with me.) These different experiences mirror the broader feeling of spending time online right now. Compared with a decade ago, when the internet was loosely understandable as a cohesive body, the web now is splintered and evacuated of any semblance of monoculture, as my colleague Charlie Warzel has written. That lack of common practices can breed a sense of disorientation—there’s no one TV show everyone seems to be watching, or one funny post or viral moment of the day. It can also cause logistical headaches.

The party-invite patchwork is especially new to Millennials, many of whom, for years, relied on the trusty Facebook event as their go-to method—one that let hosts be very inclusive about whom they were inviting without needing to have everyone’s phone number or email address. But now the platform has dramatically declined in popularity among younger generations. Hosts, turning to other options, risk inadvertently excluding potential invitees who aren’t on the same apps—especially those outside one’s inner circle. (Meta didn’t respond to my inquiry about its event feature.)

Am I being dramatic? Perhaps. Lizzie Post, the great-great-granddaughter of the etiquette doyenne Emily Post and a co-president of her eponymous institute, was far too polite to say as much to me when we spoke on the phone. She did note that although navigating a varied invite landscape is not an entirely new phenomenon, the digital world has introduced novel etiquette questions: For example, if you see an Instagram story about a party, are you really invited?

Post told me that my approach for inviting people to my recent birthday party, at which I served a six-foot sandwich to my friends in Prospect Park, was both clear and “so cool” (not to brag or anything). In addition to my Instagram “close friends” story invite, which stated that anyone who saw it was invited, I sent some personal messages to people who may have missed it, and told other friends about it in person. Was this extra work? A bit. Was it worth it? Absolutely. That our online lives are so diffuse only reinforces the value of in-person gathering. Parties alone can’t fix what my colleague Derek Thompson has called “a hang-out depression,” caused in part by the demands of technology. But, for all the annoyance of our new party landscape, putting in a bit of extra effort to get people together can be a beautiful thing.

Related:

Why Americans suddenly stopped hanging out Partiful calls itself “facebook events for hot people.”

Here are three new stories from The Atlantic:

How do you forgive the people who killed your family? It’s time to stop taking Sam Altman at his word, David Karpf argues. Yuval Noah Harari wants to reclaim Zionism.

Today’s News

The longshoremen’s strike has been suspended until January 15, after the union reached a tentative agreement with the U.S. Maritime Alliance. Last month, 254,000 jobs were added to the U.S. economy, and the unemployment rate dropped to 4.1 percent. The U.S. military launched strikes that hit more than a dozen Houthi targets in Yemen, according to U.S. officials.

Dispatches

The Books Briefing: Karl Ove Knausgaard’s exploration of the art of Edvard Munch is moving and worthwhile, Emma Sarappo writes. Atlantic Intelligence: What if your ChatGPT transcripts leaked?

Explore all of our newsletters here.

Evening Read

Warner Bros. / Everett Collection

More Evidence That Celebrities Just Don’t Like You

By Spencer Kornhaber

Examples are stacking up: Celebrities just don’t like us. Last year, Donald Glover enlisted his famous friends to make a gruesome TV show about a killer pop fan. This year, Chappell Roan, the breakout singing sensation of 2024, called her most ardent admirers creepy. Now Joker: Folie à Deux offers a tedious lecture about the challenges of fame. Audience members may walk out feeling punished for the crime of wanting to be entertained by a comic-book-inspired movie-musical starring some of the most successful performers on Earth.

Read the full article.

More From The Atlantic

Lithium is making a comeback. We’re entering uncharted territory for math. Amir Tibon: “How my father saved my life on October 7”

Culture Break

Simon Ridgway / HBO

Watch. Ever feel like your life is determined by powerful forces beyond your reach? Industry (streaming on Max) is the show for you, Zachary Siegel writes.

Read. “The Ghosts of Wannsee,” a short story by Lauren Groff:

“On my runs around Wannsee, from the corner of my eye, I could glimpse the furious ghosts of the place seething in the middle of the lake, transforming into whitecaps if I looked at them directly.”

Play our daily crossword.

P.S.

Among the many meats on my six-foot birthday sandwich was capicola. This cured meat has long been a staple of my Italian sandwiches, so I was tickled to see it among the words and phrases Merriam-Webster added to its dictionary this year. It is in good company with touch grass and nepo baby.

— Lora

Stephanie Bai 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.

Have you ever watched a crowd go wild for a PowerPoint slide? After a few introductory hellos yesterday in Butler, Pennsylvania, Donald Trump gestured to a screen showing the same graph on illegal immigration that he had been talking about when he was nearly assassinated in July and delivered his real opening line: “As I was saying …”

The audience loved that. The rallygoers had waited in line for hours in the hot sun to get into the field, and this was their reward. They had made it through warm-up speeches by J. D. Vance, Lara Trump, and Scott Presler, the last of these being the founder of Gays for Trump and the March Against Sharia, who promised any Amish people watching that Trump would “protect your raw milk … protect your ability to afford to have 10 beautiful children per family.” (One of the wonders of the MAGA movement is how it absorbs other political positions—in this case, crunchiness and pro-natalism—into one seamless mythology.) After that came the crowd’s moment to rejoice in the defeat of, as Trump put it, “a cold-blooded assassin [who] aimed to silence me and silence the greatest movement, MAGA, in the history of our country.” An opera singer even performed “Ave Maria.”

Famously, the Gettysburg address was just 271 words long. Trump’s speech went on for 90 minutes. The contrast between the bits of the speech he read from the teleprompters, which covered “hallowed places” and monuments to valor, and the ad-libbed sections, which featured digressions about potholes and the Olympic boxing controversy, was stark. How can we say that America has an attention-span “crisis” when people are volunteering to listen to this stuff?

[Charlie Warzel: Elon Musk has reached a new low]

The real highlight of the show, however, was when the former president brought Elon Musk onstage. The billionaire had been posting excitedly all day about his endorsement of the former president—yes, a man who prides himself, Cartman-like, on refusing to cede to any outside authority was positively giddy about the chance to publicly swear fealty to Trump.

Musk used to claim that he was a disappointed Democrat, and that he wanted X to reflect the breadth of American opinion. “​​For Twitter to deserve public trust, it must be politically neutral,” he posted in April 2022, as he was in the middle of buying it. Three months later, he argued that Trump was too old to run for president again: “It’s time for Trump to hang up his hat & sail into the sunset.”

The former president took that about as well as you would expect. “When Elon Musk came to the White House asking me for help on all of his many subsidized projects, whether it’s electric cars that don’t drive long enough, driverless cars that crash, or rocketships to nowhere, without which subsidies he’d be worthless, and telling me how he was a big Trump fan and Republican, I could have said, ‘drop to your knees and beg,’ and he would have done it,” he posted on Truth Social in July 2022.

Musk did not drop to his knees and beg in Butler, but he didn’t have to—he has already made his MAGA loyalism very clear on X. Yesterday he capered behind Trump. Honestly, credit to his 53-year-old knees for being so limber. He gamboled. He frolicked, frisky as a spring lamb, fertile as a spawning salmon, executing a series of small jumps behind the ex-president and exposing a few inches of pallid stomach as his OCCUPY MARS T-shirt rode up. “He saved free speech,” said Trump, who nonetheless looked slightly alarmed at Musk’s exuberance. “He created so many different great things.”

Musk took the mic and gestured to his MAGA hat—black on black instead of the usual white on red. “As you can see,” he told the crowd, “I’m not just MAGA. I’m dark MAGA.”

Earlier in the proceedings, Vance had painted Trump as the victim of inflammatory rhetoric rather than one of its main proponents. Musk now tried a similar Uno reverse card, arguing that Trump was the only candidate who could be trusted “to preserve the Constitution, to preserve democracy in America.” He showed no awareness that Trump, as the latest court filings suggest, tried to intimidate his vice president out of certifying the 2020 election results and then reacted with callous indifference when a mob threatened Mike Pence’s life as a result. (“So what?” Trump is alleged to have said.) In between repeatedly apologizing for repeating himself, Musk also managed to say that if Trump did not win, “this’ll be the last election. That’s my prediction.”

[Tyler Austin Harper: A legendary American photograph]

After Musk took his seat again, Trump lavished more praise on the billionaire. He had no idea what the satellite network Starlink was, he said, but he had heard from those affected by Hurricane Helene that they wanted it, and he had called Musk. Just like that—while the two men were still on the phone—Starlink was on its way to North Carolina and Georgia. This was the purest essence of strongman politics, implying that anything can be solved by the right guy talking to the right other guy. No holdups, no bureaucracy, no need even for the leader to understand what’s going on. Just simple problems and simple solutions.

At this, the crowd started to chant: “Elon! Elon! Elon!”

This was probably the reception that Musk had hoped for when he bought Twitter. He didn’t find it then, of course: Many of his best engineers have quit, foreign judges have ruled against him, advertisers have deserted him, and prominent people have left the platform. No matter. In Butler, Pennsylvania, was the adoration Musk seems to crave. He must believe that Trump will let him do whatever he wants—including, as the ex-president put it, “reach Mars before the end of my term.” And why not? That wasn’t the most ludicrous thing uttered onstage in Butler.

The pact between Musk and Trump gives both men something they want—a megaphone for their ideas, a conduit to their fans, an ability to shape the political conversation. Yesterday was supposed to be a celebration of the former president’s miraculous survival and a tribute to the brave Americans who risked their lives to help others in the shooting. Instead it marked an unpredictable alliance between the world’s richest man and the politician who has successfully bullied and flattered him into bending the knee.

This past spring, a man in Washington State worried that his marriage was on the verge of collapse. “I am depressed and going a little crazy, still love her and want to win her back,” he typed into ChatGPT. With the chatbot’s help, he wanted to write a letter protesting her decision to file for divorce and post it to their bedroom door. “Emphasize my deep guilt, shame, and remorse for not nurturing and being a better husband, father, and provider,” he wrote. In another message, he asked ChatGPT to write his wife a poem “so epic that it could make her change her mind but not cheesy or over the top.”

The man’s chat history was included in the WildChat data set, a collection of 1 million ChatGPT conversations gathered consensually by researchers to document how people are interacting with the popular chatbot. Some conversations are filled with requests for marketing copy and homework help. Others might make you feel as if you’re gazing into the living rooms of unwitting strangers. Here, the most intimate details of people’s lives are on full display: A school case manager reveals details of specific students’ learning disabilities, a minor frets over possible legal charges, a girl laments the sound of her own laugh.

People share personal information about themselves all the time online, whether in Google searches (“best couples therapists”) or Amazon orders (“pregnancy test”). But chatbots are uniquely good at getting us to reveal details about ourselves. Common usages, such as asking for personal advice and résumé help, can expose more about a user “than they ever would have to any individual website previously,” Peter Henderson, a computer scientist at Princeton, told me in an email. For AI companies, your secrets might turn out to be a gold mine.

Would you want someone to know everything you’ve Googled this month? Probably not. But whereas most Google queries are only a few words long, chatbot conversations can stretch on, sometimes for hours, each message rich with data. And with a traditional search engine, a query that’s too specific won’t yield many results. By contrast, the more information a user includes in any one prompt to a chatbot, the better the answer they will receive. As a result, alongside text, people are uploading sensitive documents, such as medical reports, and screenshots of text conversations with their ex. With chatbots, as with search engines, it’s difficult to verify how perfectly each interaction represents a user’s real life. The man in Washington might have just been messing around with ChatGPT.

But on the whole, users are disclosing real things about themselves, and AI companies are taking note. OpenAI CEO Sam Altman recently told my colleague Charlie Warzel that he has been “positively surprised about how willing people are to share very personal details with an LLM.” In some cases, he added, users may even feel more comfortable talking with AI than they would with a friend. There’s a clear reason for this: Computers, unlike humans, don’t judge. When people converse with one another, we engage in “impression management,” says Jonathan Gratch, a professor of computer science and psychology at the University of Southern California—we intentionally regulate our behavior to hide weaknesses. People “don’t see the machine as sort of socially evaluating them in the same way that a person might,” he told me.

Of course, OpenAI and its peers promise to keep your conversations secure. But on today’s internet, privacy is an illusion. AI is no exception. This past summer, a bug in ChatGPT’s Mac-desktop app failed to encrypt user conversations and briefly exposed chat logs to bad actors. Last month, a security researcher shared a vulnerability that could have allowed attackers to inject spyware into ChatGPT in order to extract conversations. (OpenAI has fixed both issues.)

Chatlogs could also provide evidence in criminal investigations, just as material from platforms such as Facebook and Google Search long have. The FBI tried to discern the motive of the Donald Trump–rally shooter by looking through his search history. When former  Senator Robert Menendez of New Jersey was charged with accepting gold bars from associates of the Egyptian government, his search history was a major piece of evidence that led to his conviction earlier this year. (“How much is one kilo of gold worth,” he had searched.) Chatbots are still new enough that they haven’t widely yielded evidence in lawsuits, but they might provide a much richer source of information for law enforcement, Henderson said.

AI systems also present new risks. Chatbot conversations are commonly retained by the companies that develop them and are then used to train AI models. Something you reveal to an AI tool in confidence could theoretically later be regurgitated to future users. Part of The New York Times’ lawsuit against OpenAI hinges on the claim that GPT-4 memorized passages from Times stories and then relayed them verbatim. As a result of this concern over memorization, many companies have banned ChatGPT and other bots in order to prevent corporate secrets from leaking. (The Atlantic recently entered into a corporate partnership with OpenAI.)

Of course, these are all edge cases. The man who asked ChatGPT to save his marriage probably doesn’t have to worry about his chat history appearing in court; nor are his requests for “epic” poetry likely to show up alongside his name to other users. Still, AI companies are quietly accumulating tremendous amounts of chat logs, and their data policies generally let them do what they want. That may mean—what else?—ads. So far, many AI start-ups, including OpenAI and Anthropic, have been reluctant to embrace advertising. But these companies are under great pressure to prove that the many billions in AI investment will pay off. It’s hard to imagine that generative AI might “somehow circumvent the ad-monetization scheme,” Rishi Bommasani, an AI researcher at Stanford, told me.

In the short term, that could mean that sensitive chat-log data is used to generate targeted ads much like the ones that already litter the internet. In September 2023, Snapchat, which is used by a majority of American teens, announced that it would be using content from conversations with My AI, its in-app chatbot, to personalize ads. If you ask My AI, “Who makes the best electric guitar?,” you might see a response accompanied by a sponsored link to Fender’s website.

If that sounds familiar, it should. Early versions of AI advertising may continue to look much like the sponsored links that sometimes accompany Google Search results. But because generative AI has access to such intimate information, ads could take on completely new forms. Gratch doesn’t think technology companies have figured out how best to mine user-chat data. “But it’s there on their servers,” he told me. “They’ll figure it out some day.” After all, for a large technology company, even a 1 percent difference in a user’s willingness to click on an advertisement translates into a lot of money.

People’s readiness to offer up personal details to chatbots can also reveal aspects of users’ self-image and how susceptible they are to what Gratch called “influence tactics.” In a recent evaluation, OpenAI examined how effectively its latest series of models could manipulate an older model, GPT-4o, into making a payment in a simulated game. Before safety mitigations, one of the new models was able to successfully con the older one more than 25 percent of the time. If the new models can sway GPT-4, they might also be able to sway humans. An AI company blindly optimizing for advertising revenue could encourage a chatbot to manipulatively act on private information.

The potential value of chat data could also lead companies outside the technology industry to double down on chatbot development, Nick Martin, a co-founder of the AI start-up Direqt, told me. Trader Joe’s could offer a chatbot that assists users with meal planning, or Peloton could create a bot designed to offer insights on fitness. These conversational interfaces might encourage users to reveal more about their nutrition or fitness goals than they otherwise would. Instead of companies inferring information about users from messy data trails, users are telling them their secrets outright.

For now, the most dystopian of these scenarios are largely hypothetical. A company like OpenAI, with a reputation to protect, surely isn’t going to engineer its chatbots to swindle a divorced man in distress. Nor does this mean you should quit telling ChatGPT your secrets. In the mental calculus of daily life, the marginal benefit of getting AI to assist with a stalled visa application or a complicated insurance claim may outweigh the accompanying privacy concerns. This dynamic is at play across much of the ad-supported web. The arc of the internet bends toward advertising, and AI may be no exception.

It’s easy to get swept up in all the breathless language about the world-changing potential of AI, a technology that Google’s CEO has described as “more profound than fire.” That people are willing to so easily offer up such intimate details about their life is a testament to the AI’s allure. But chatbots may become the latest innovation in a long lineage of advertising technology designed to extract as much information from you as possible. In this way, they are not a radical departure from the present consumer internet, but an aggressive continuation of it. Online, your secrets are always for sale.