The book currently topping the New York Times nonfiction bestseller list is The Anxious Generation, a jeremiad against social media and its impact on young people. Its thesis is that apps like Facebook, Instagram, and TikTok have unleashed an epidemic of mental illness among children, preteens, and teenagers. Immediate and extreme measures are required to reverse this deadly trend. The author, NYU social psychologist Jonathan Haidt, has been promoting it everywhere. Parents have rallied around his thesis, and Haidt’s claims have thrown gasoline on a smoldering movement to pass new laws to limit social media. But a review in Nature, one of the world’s most prestigious scientific journals, hit a more critical note. “The book’s repeated suggestion that digital technologies are rewiring our children’s brain and causing an epidemic of mental illness,” wrote Candice L. Odgers, a professor of psychological science at UC Irvine, “is not supported by science.”
This academic tiff has real consequences—think of the kids! When I contacted Odgers for more background, she was happy to oblige. She and her academic peers seem to view the massive public embrace of Haidt’s book as a slow-motion horror show. “I’ve been researching adolescent mental health for 20 years, and then actually tracking young adolescents, 10 to 14 years old, on their phones since 2008,” she says. “So I'm kind of old in this game. He's new.” Odgers says that the consensus of the community that actually studies this stuff for a living is that there is no evidence of Haidt’s claims that social media has triggered a mental health epidemic among a generation whose brains were scrambled by swiping on smartphones. She pointed me to a number of studies, including one 2022 meta-analysis of 226 studies involving 275,728 participants. “The association between social media use and well-being was indistinguishable from zero,” she and her coauthor reported. A study completed last year by the National Academy of Sciences concluded there was not enough evidence to link social media with changes in adolescent health. (It asks for more research, measures to minimize any harm by social media, and also steps to maximize its benefit.)
Odgers and a colleague did find some studies that indicated there might be health effects of social media, but when they looked into it those experiments weren’t necessarily directed at kids. “Participants were often middle-aged women recruited online or small samples of college students who were asked to give up social media and report how they feel,” Odgers and her collaborator wrote in an article entitled “Let’s Stop Shaming Teens About Social Media Use.” Her bottom line is that while there is indeed a troubling rise in mental health issues among young people, especially girls, nothing beyond a nostalgia-tinged gut reaction indicates that social media is to blame.
When I laid this out to Haidt himself in a phone interview, he was well prepared. He describes the pushback against his work as “a normal academic dispute.” To his credit he has engaged with some of the criticism in his Substack posts. He’s not planning a retreat from his thesis. “I’m not going to convince them, and they're not going to convince me,” he says. “We each make our best arguments, and then the rest of the academic community will tune in and decide if there’s evidence of harm here.”
One of Haidt's strongest arguments is that it seems obvious how social media would play a role in the rise in mental health issues, even if other causes are just as likely. “We can say with confidence that teenagers are susceptible to the moods and beliefs of other teenagers—it’s been true forever,” he says. “We can also say that girls are more so than boys; they share each other's emotions more. So the idea that girls are picking up depression and anxiety from other girls I think is not controversial. The question is: What else are they picking up? And we don't know. Nobody knows.”
It would help if we did know. This argument matters in part because legislators and regulators in the US, EU, and UK are considering restrictions on how—and whether—minors access social media. It would make some sense if the threat was determined or debated on solid scientific ground. In the US, the leading potential bill is The Kids Online Safety Act, (Kosa). A majority of the Senate has endorsed this draft law, but some civil liberties and free speech organizations think the bill goes too far. After objections from LGBTQ+ organizations that the bill would deny young people access to resources, its authors made rewrites to mitigate that danger. But critics feel that the bill still will provide too much authority for state and local officials to censor content on social media in the name of protecting children. A different bill would outright ban social media to everyone under 13, though some worry that if it were passed, stealthy apps would appear that offer even fewer protections than mainstream ones do now.
Putting the correlation-versus-causation issue aside, some of Haidt’s critiques of social media directed at kids seem beyond dispute. Meta’s own studies show that Instagram can have a negative effect on teen girls. It’s also clear that companies like Meta—which owns Instagram and Facebook—have not done enough to protect kids. During Mark Zuckerberg’s painful appearance at a Senate hearing last January—the one where the Meta CEO choked out regret to parents whose kids committed suicide after interactions on his platform—Senator Richard Blumenthal revealed a damning chain of 2021 emails. Meta’s top executives, including then-vice president of global affairs Nick Clegg, chief product officer Chris Cox, and then-chief operating officer Sheryl Sandberg, pleaded for more resources to address child safety and well being, laying out detailed plans that included more employees working on the problem. But Zuckerberg nixed the plan, basically because he didn’t want to spend the money. (Meta spokesperson Andy Stone says that the company has developed more than 50 tools to help teens have safe experiences.)
But while Haidt harshly criticizes the inaction and greed of tech companies, he places most of the onus on the technology itself. He classifies social media as inherently evil like booze, cigarettes, and automatic rifles. That’s a harsh conclusion, especially if there’s only shaky evidence, or no evidence at all, that those apps are the key driver in kids’ mental health struggles. Especially since, unlike nicotine, liquor, and AR-15s, those apps have tangible benefits to young people.
Yes, some kids suffer because social media exposes them to bullies, makes them hate their bodies, or exposes them to unhealthy advertising. But one might argue it’s easier to find evidence that social media in the aggregate benefits youths more than it harms them. A sweeping US Centers for Disease Control study found that during the pandemic, there was indeed a mental health crisis among adolescents, and alarming rates of suicide. But social media could sometimes mitigate those problems. The study found that the bullying that kids suffered was less likely coming from their peers on Instagram than within their own families. And social media actually helped those kids: “Youth who felt more connected to people at their schools had better mental health,” wrote the CDC in a summary page, in a section called “Connectedness Protects Youth.”
Haidt gives short shrift to the idea that social media has benefits. Intentionally or not, his arguments could lead to kids getting less useful information relevant to their lives. “He generally centers technology as the greatest cause of all the problems of young people, which is not accurate,” said technosocial scholar danah boyd, founder of the Data & Society research nonprofit, on a recent podcast. “It’s a really successful thing because it feeds into parents’ anxieties.”
Taylor Lorenz, the Washington Post reporter covering social media and author of the recent book Extremely Online, says that Haidt’s arguments are toxic. “It’s harmful because, as so many LGBTQ+ groups and others have made it very clear, social media is a lifeline for millions and millions of people,” she says. “And the notion that screen time is inherently bad is also a regressive idea. The same arguments that he uses are the ones that people once used for banning comic books, the radio, and all new forms of media.”
Haidt himself does say that social media isn’t the only thing that thwarts normal development for young people. The other trend among the young he bemoans is how overprotective parenting has cut down on IRL interactions, a phenomenon he addressed in his previous book, The Coddling of the American Mind. At one point in his new book, Haidt expresses unhappiness that hospitals aren’t seeing as many boys show up with shattered limbs, because it means that young males aren’t bulking up with the requisite toughness that will stand them in good stead during adulthood. He seems to idealize the Leave It to Beaver-style aimless hangouts of an earlier era, taking for granted that it’s a superior environment than the online one where people might make friends from anywhere in the world but not fall out of trees with them. I’m not so sure: I grew up way before social media and can attest that bullying in the real world was pervasive long before the internet showed up. At least on the internet, you don’t get punched in the face.
Like it or not, being digitally connected is the reality of the 21st century. This clock can’t be rolled back, and whatever comes after smartphones is going to make us even more connected. It’s unrealistic to treat digital life—aka just life—like an NC-17 movie that screens kids at the box office. Of course, we should demand that children be safe in this realm, as in any other, and I agree with Haidt that it’s a good idea to ban phones during school hours. But something big is afoot that affects all of humanity, and we have voted with our finger-swipes that it’s something we want.
“We're not going to stop the great rewiring of humanity,” Haidt concedes when I present him with that argument. “But it's an open question: Is this harmful to children?” The scientific community appears to be less sure, given the evidence cited by Odgers suggesting that the mental health crisis among kids has other causes. And for Haidt, the question seems already closed—his book screams that kids are being ruined. “I see no sign that kids can make it through this rewiring when we take away normal human childhood and give them a rewired child,” he tells me. He’d better look harder for those signs, because in 2024, “a normal childhood” now involves technology and social media. We could and should make it safer and less exploitative. We must demand that the platforms do better. But there’s no walking back the march of digital connection.
Time Travel
Almost 30 years ago, as the internet was finding its way into our lives, the media started writing stories about “internet addiction.” Like Jonathan Haidt’s current campaign against social media, it put a scientific sheen on some actual problem exacerbated by technology, while deemphasizing the value people extracted from online activity. I wrote about it for Newsweek in December 1996, in an article headlined “Breathing Is Also Addictive.”
The article in The New York Times painted a horrifying picture. A 17-year-old boy in a Texas rehab center, alternatively convulsing and throwing furniture around the room. “Had he been hooked on heroin?” the writer breathlessly asked. “Cocaine? Jim Beam? Joe Camel?” No. This was a victim of the latest dark cohort of that addictive cabal: the Internet. Another poor bastard hooked on cyberspace. And he's just the tip of the iceberg, the Times told us. Watch out for an epidemic of Net junkies. Moms who Net-surf while doe-eyed babes go hungry. Geeky high-school students up until 4 a.m., assuming studly personae in multiplayer dungeon games. Broken homes, broken jobs, broken lives ... the Days of Wine and Modems! All because of that Demon Net.
The problem with that article, and dozens like it that ran in 1996, is that "Internet addiction'' doesn't really exist. This disconnect between reality and reporting was symbolic of the general media's inability to convey the boggling changes that society is about to undergo, all because of our impending connectedness. The media declared 1995 the Year of the Internet; in '96, we tried to explain it to you. The measure of our success, sadly, may one day be judged by the Internet-addiction scare.
For starters, the term was coined as a joke by New York City psychiatrist Ivan Goldberg, who was astonished that people took him seriously. “There ain't no such thing!” he would howl to journalists, who wrote their stories anyway. Yes, it is true that many people use the Internet for hours on end; and, yes, it is true that a small percentage of those people go overboard, preferring the companionship of their online buddies to whining kids, demanding spouses and tedious responsibilities in the so-called “meat world.”
But these people aren't addicted, certainly not in the way that alcohol, drugs or cigarettes cause a measurably physical craving. Nor can you compare the Net's lure to gambling and overeating fixations: those are destructive activities, and the Net is not. “Using the Internet,” writes psychologist Storm A. King, “is no more inherently addicting than any other human activity that someone might find pleasure-producing, valuable, or productive.”
Ask Me One Thing
Tom asks, “The internet is outraged about a new Apple ad called “Crush!” in which musical instruments and other creative objects and materials are squashed by a giant machine and emerge as a brand-new iPad Pro. What do you think of it?”
Thanks, Tom. Apple has been a laggard among elite tech companies in generative AI. Though artificial intelligence has long been baked into Apple’s products and services, the company’s dramatic full-body embrace into tech’s most vital and disruptive sector apparently won't happen until next month’s developer conference, WWDC. This has somewhat insulated Apple from the growing unease about how AI will dehumanize our products, services, jobs, and even humans. As a result, it unwittingly made a one-minute video that triggered an artistic community already outraged by AI’s incursion on its turf. Microsoft, OpenAI, or Google never would have made this ad, because its leaders would have understood the implications in a second.
That’s the only explanation I can see for an ad that fetishizes the destruction of beloved instruments of human creativity. For those who haven’t viewed it, the commercial shows an industrial crush machine flattening musical instruments, video consoles, televisions, turntables, books, paints, and other beloved items with the enthusiasm of an action film glamorizing a car wreck. It’s like a snuff film where art itself is lasciviously murdered. The ostensible message is that all the wonderful songs, movies, video games, and books created with the artifacts of analog are now preserved in a gorgeous sliver of glass and aluminum of the new iPad Pro, which is all that remains when the brutal pancaking is complete. But that signal is overwhelmed by the prurient noise of destruction.
Instead of an ad for a cool new hardware product, Apple has minted a visible parable embodying the fears that AI will eliminate humanity’s purest expressions. It’s easily the biggest Apple ad blunder since “Lemmings” in 1985, which idiotically attempted to woo IBM customers by implying they were dull, brainsnatched droids. It’s also traumatic to inflict this carnage on a public that has yet to process the South Dakota governor’s boast that she murdered a 14-month-old puppy. Watching an acoustic guitar being crushed is sort of equivalent when it comes to fears that AI will supplant artists.
But here’s evidence that creativity still rules—one clever video artist fixed the problem by running the ad in reverse. Instead of unintentionally proclaiming the end of human creativity, the rewind version starts with the iPad and shows how within its wispy profile all the tools of genius reside therein. Tim Cook would do well to reward that creator and run the reverse version of the ad immediately in the original’s place. One also hopes this experience will incentivize Apple to approach the age of AI with more humility. We’ll find out next month.
You can submit questions to mail@wired.com. Write ASK LEVY in the subject line.
End Times Chronicle
Tornado-mania breaks out—just as the cicada invasion begins, with Illinois the epicenter of an unholy merger of two massive and destructive breeds. Who said life is boring between the coasts?
Last but Not Least
Elon Musk’s Neuralink’s technology is still in the brain of its first experimental patient. But not all is well.
After solving the very hard problem of protein folding with AlphaFold, DeepMind is now using its AI technology to model DNA.
An interview with a representative of the hacking group Cyber Army of Russia was bound to be strange. And it is.
Black Twitter was glorious. Fortunately its highlights are preserved in a terrific documentary, based on WIRED reporting, that you can view on Hulu.
