College grads are rejecting AI en masse

‘Tis the season of college commencement—and this year, it’s also the time for booing commencement speakers, who have made AI their go-to topic in the face of vocal student opposition. At the University of Central Florida in May, real estate exec Gloria Caulfield was virally booed for boosting AI as a happy “next industrial revolution” for grads. At Marquette University, Adobe’s Chris Duffey was similarly disdained for an AI-happy pep talk. At Middle Tennessee State University, music executive Scott Borchetta was the one mocked for cheering on the rise of AI in the music business.

Former Google CEO Eric Schmidt was jeered at for telling the graduating class at the University of Arizona that “the question is whether you will help shape artificial intelligence.” And when current Google CEO Sundar Pichai spoke Friday at Stanford University’s graduation, he mostly avoided the topic: People had been giving him “a lot of advice,” Pichai admitted, about what not to say.

And that makes sense. After all, the class of 2026 is part of a generation facing high levels of underemployment and joblessness thanks to AI: about half of Gen Z workers believe the technology’s dangers are greater than its value, and some six in ten say they’re anxious about it.

Silicon Valley bigshots have been fixtures on the graduation circuit for decades, but they’ve traditionally been well received; for most of that time, they represented an exceptionally lucrative, growing sector that drew a massive share of young grads. Steve Jobs’ 2005 Stanford commencement speech—“Stay hungry. Stay foolish.”—remains an icon of the genre; the guy giving it was part of an elite that the young listener could conceivably join.

The intense contempt for AI at college graduations is a glaring sign of how many of the young are AI refuseniks—and that tech representatives now look hungry, foolish, and unable to read the room.

Zeynep Maya Dogancioglu, a University of Tampa student who leads the local chapter of a group called Pause AI US, tells me that she thinks students are “completely right” to heckle the “hypocrisy” of the pro-AI crowd.

“In every single class, they start by telling us we cannot use AI for academic integrity,” Dogancioglu said—but “the teachers who are grading us are using these same tools by choice, and the administration is pushing AI as this next big thing.”

“They are putting the wants and needs of billionaires over us,” a displeased University of Arizona student remarked of Schmidt’s pro-AI speech in the Guardian. A Duke University student pled in the campus newspaper for Zohran Mamdani to speak instead. At the University of California, Santa Barbara, Nickolas Spiliotopoulos told NBC News that students like him “don’t want AI to trump our academic, maybe our political, maybe our cognitive processes.”

To these young people, the commencement speakers are representatives of a dumbed-down, even greedier elite—a class that embodies what you could call the banality of venal.

Seventy years ago, the renowned sociologist C. Wright Mills first published The Power Elite, now a fundamental text of 20th-century political sociology, mapping the interlocking circles of the American ruling class: corporate barons, the “war lords” at the Pentagon, and political leaders. The theory was structured on Franz Neumann’s Behemoth, published in 1942, about the “structure and practice” of German’s Nazi Party—and yet it’s hard to reread Mills today without concluding that things are now worse.

Among Mills’ key observations was the rise to power, as historian and Mills specialist John Summers puts it to me, of “managers of the bureaucracy” who were “charged with solving technical problems.” For the most part, Mills found, the elite of the mid-20th century didn’t reach their positions due to great talent, intelligence or skill, but through networks of family wealth and a drive for status.

Yet they still held a modicum of public trust, particularly the administrators and scholars among them. It was a time of general prosperity, which new technology facilitated; the elite who managed those new tools were getting more for themselves, sure, but they wanted stability and predictability so they might thrive. (Indeed, they often made power seem relatively boring and procedural.)

Their successors are marked by being “disinhibited”, says sociologist Heather Gautney, whose The New Power Elite follows on Mills’ original: far less able to recognize the needs of the people whose futures they are supposed to help steward, and less likely to even gesture at sympathy for them, engaging in constant self-dealing with far less fear of exposure.

They are often more antigovernment than in the past, leaning “deeper and deeper into corporate authoritarianism,” as Gautney puts it, and preoccupied with dismantling the administrative state their predecessors ran. While Mills wrote that “great corporations are the great units of wealth, to which individuals of property are variously attached,” now, the companies are often secondary to people like Musk, who control massive firms in different sectors. The new power elite better resembles a cult structure, in Gautney’s words—as recent revelations in Wired about Dialog, Peter Thiel’s secretive elite gathering, drive home.

The new power elite despise security and continuity: they’re hell-bent on removing the bureaucracy of Mills’ era—which no one exemplifies better than Elon Musk, with the recent Texas outbreak of flesh-eating screwworm potentially linked to DOGE’s elimination of USAID. Their bright ideas include the end of liberal arts university education, as promoted by Thiel or his Palantir minion Alex Karp; it can be replaced by rented time in AI datacenters, where people will “buy” intelligence “from us on a meter,” as OpenAI boss Sam Altman has suggested.

This inability to absorb the life experience of everyday people—many of us simply want to go to college! Many of us enjoy meaningful work!—stands in sharp contrast to the behavior of the old power elite, who oversaw an epochal expansion of higher education. And as exploitative as corporations were back then, they didn’t advertise quite as cheerfully that they’d be hollowing out your industry sometime in the next quarter. 

What do they have to fear? David Ellison and his father, Larry, control Paramount and CBS; Elon Musk has turned Twitter into a right-wing slop factory; conservative Sinclair Media owns legacy papers and close to 200 television stations. The worst of the new power elite can now get away with whatever they like.

But the students’ vocal rejection of AI exemplifies what Mills called “the sociological imagination” at work. Mills’ meaning was that the public should use our interpretative gifts to connect daily life to broader social structures—and critique those who fail to do so.

Their public opposition is driving powerful figures like Google’s Pichai away from AI boosterism—and inspiring unexpected older supporters like the fashion designer Jeremy Scott, known for his high-priced ironic menswear and camp appearances on fashion-reality shows. Young people, Scott recently told the Kansas City Art Institute’s graduating class, should reject LLMs and instead embrace what they already have: “actual intelligence.”

This article was produced in collaboration with the Economic Hardship Reporting Project, which supports independent journalists as they forward fresh narratives about inequality. Subscribe to follow EHRP.