A Seat on the Rocket Ship
The students who booed AI at graduation were right. It probably won't matter.
My favorite new trend is students booing commencement speakers who try to push AI.
It started at the University of Central Florida, on May 8, at the ceremony for the College of Arts and Humanities and the Nicholson School of Communication and Media. The speaker was Gloria Caulfield, vice president of strategic alliances at Tavistock Development Company, a Florida real estate firm. She’d been going for a few minutes, working through the usual commencement throat-clearing about how change is scary but also exciting, when she arrived at this: “The rise of artificial intelligence is the next industrial revolution.”
The room booed her.
Not a scattered grumble. A real, rolling wave of it, loud enough that she stopped, turned around, and stood there with her hands out, trying to work out what had gone wrong. “Okay, I struck a chord,” she said. “May I finish?” Somewhere in the crowd, a graduate yelled “AI sucks.” She pushed on, reached the line “only a few years ago, AI was not a factor in our lives,” and the room cheered. They were cheering, at a graduation, for the memory of a world that existed before ChatGPT. Caulfield, rattled now, called it “a bipolar topic,” and a minute later, getting booed again, settled on a reading of the moment that let her keep her feet: “I love it. Passion, let’s go.”
That weekend, it was Scott Borchetta’s turn. Borchetta founded Big Machine Records, the label that put out Taylor Swift’s early albums, and he was the commencement speaker at Middle Tennessee State University, where some of the graduates in front of him were collecting degrees from its Scott Borchetta College of Media and Entertainment. He told them “AI is rewriting production as we sit here,” and the room booed. Borchetta’s answer was not Caulfield’s. “Deal with it,” he said. “It’s a tool. Make it work for you.” To the people still booing him, he added: “Hear me now or pay me later.”
Then, the next week, it happened again! Eric Schmidt, the former Google CEO, gave the commencement address at the University of Arizona on May 15, and the graduates booed him too, loudest every time he turned to AI. Arizona is the messy example of the three. Students there had spent weeks organizing against Schmidt’s appearance over a sexual assault lawsuit filed by his former partner, and the booing started well before he got to AI at all. UCF and MTSU are cleaner. But the throughline holds across all of them: stand up in front of a graduating class in 2026, tell them their future runs on AI, and a good chunk of the room will let you know exactly what they think of that.
There’s something clarifying about a boo! It’s not a subtle instrument. Nobody has ever had to wonder what a few thousand people booing in unison were trying to get across. A boo has nothing in it to misread, no second meaning tucked behind the first. It isn’t really language. It’s volume aimed at a person, and it means no.
So let’s talk about it.
The students aren’t confused
The simplest part first: the students were right. Being right here didn’t take any special insight. They are about to graduate into a labor market that sucks and has been getting worse for years for people exactly like them, and they have noticed.
Madison Fuentes graduated from UCF this month with a degree in creative writing, and she told the Orlando station News 6 what was behind the booing: “I don’t think that kids are having a hard time accepting it because we know that AI exists. I think we’re just having a hard time acknowledging that it’s taking away job opportunities from us.” Fuentes and her classmates use AI. They’ve used it for years; about half of Gen Z uses generative AI at least weekly. The booing came from people who know the technology well and have done the math on what it means for them. The numbers are on their side, too. The education company Cengage found that just 30% of last year’s graduates landed full-time work, down from 41% the year before.
And the people building AI know all of this.
In April, the New York Times ran a piece by Jasmine Sun headlined “Silicon Valley Is Bracing for a Permanent Underclass,” about the mood inside the industry. Sun’s sources, she wrote, were a lot grimmer about AI and jobs in private than they were on the record. They “suddenly became optimists once I turned on the mic.”
While I was working on this, Brian Merchant published a piece at his excellent newsletter, Blood in the Machine (named after his book on the Luddites), that opens on the very same UCF clip. His argument, in short, is that the anger at AI is mostly anger at capitalism: AI has become the most visible face of an economy that was already stacked against the people now graduating into it, and the students booing understand that just fine. He makes that case better than I’m going to, and if it’s the case you came for, go read him. What I want to do here is follow the boo a little further, out of the auditorium and into the hands of the people who decided what it meant.
The room read her fine
Some of the write-ups of the Schmidt speech had him failing to “read room on AI.” Orlando Weekly called Caulfield’s line “ill-timed.” Plenty of the coverage was sharper than that, but the reflex was consistent: the booing was a misjudgment, a speaker who’d gotten her audience wrong.
I think that phrase is doing more work than it looks like.
I’d argue that “read the room” is a phrase mostly about delivery. It assumes the speech itself was basically fine and the speaker just aimed it badly, at the wrong crowd on the wrong day. Get the aim right and the speech lands. That’s the entire theory of the phrase. So does it hold up here? Is there a version of “the rise of artificial intelligence is the next industrial revolution” that a room of about-to-graduate humanities majors stands up and cheers for? Maybe. What’s the line a better-calibrated Caulfield delivers instead?
I don’t think there is one, because there’s really no gentle way to tell a graduating class that the economy is being rebuilt to need fewer of them. They heard the content. They heard it correctly. The delivery was never the problem.
And the boos were serious, and specific. Watch the rest of Caulfield’s speech and you find that her actual closing message, the inspirational note she chose to leave a room of arts and communications graduates with, was that “communication is a superpower.” Her example was Magic Johnson, and the story of how he once talked Starbucks CEO Howard Schultz into a business partnership. That was the send-off: here is how a person persuades a CEO. Houda Eletr, a UCF graduate quoted by Orlando Weekly, called Caulfield a “corporate mouthpiece.” Oof!
When “read the room” runs out, there’s always “Luddite,” a word that pulls the same trick: it turns a person with an objection into a person who’s just scared of machines. (Read Merchant’s book to better understand who the Luddites actually were, please.) Either way, the move is the same. Make the booing a feeling, a failure of manners or nerve, anything but a judgment.
Don’t worry about it
It was interesting to watch how the people actually being booed responded to it live, while it was still happening to them. Mostly, what they did was reassure the room. (Borchetta, telling a booing crowd to “deal with it,” was the blunt exception.) And the AI industry keeps two reassurances on hand.
Schmidt’s was the more sophisticated one. He didn’t pretend the fear wasn’t there. He named it, and named it accurately: he told the Arizona graduates he could hear them, that they were afraid the machines were coming and the jobs “evaporating.” And he agreed the fear was “rational.” Then, in the same breath, he told them where it comes from. The fear, he said, is “amplified every day by social media platforms with algorithms that have learned with great precision that fear earns clicks and that anxiety drives engagement.”
A man who ran Google for a decade told a stadium of young people that their dread about the tech industry is real, then placed the source of that dread inside their own phones.
The second reassurance is for when the first one won’t hold, when the job losses get too concrete to wave off as a feeling. Then the answer becomes: don’t worry, there will be a safety net. And the safety net the industry reaches for, every time, is universal basic income.
UBI is a strange thing to keep promising, because it is close to the least likely policy in American life to actually happen. This is a country whose politics include a durable, bipartisan enthusiasm for drug-testing welfare recipients and arguing over whether food stamps should be allowed to cover soda. UBI gets named the way you make a vague promise to a kid who won’t stop asking. It’s there to close the subject, and most of the time it works.
The industry’s own behavior gives the game away. Sun’s piece describes the Democratic pollster David Shor laying out polling for a room of AI workers: universal basic income tests badly, while a federal jobs guarantee is much more likely. A safety-net policy that could actually pass exists; it just isn’t the one the industry names. Part of that, I believe, is because once you start talking about a jobs guarantee, you have to start addressing what these jobs look like.
The two reassurances don’t fit together, and the contradiction is total. One treats the fear as a glitch your feeds installed in you. The other treats the fear as accurate, and promises a fix. The industry never reconciles them. It just reaches for whichever one the moment calls for. Either way, the boo gets answered without being answered.
Then it became a word
When the AI backlash throws off another clip, another viral moment, the people whose job is to explain things go to work. I’m one of them. And one of the moves the job reaches for, again and again, is to give the thing a name.
The name that’s gotten attached to it is “AI populism.” The phrase itself isn’t new, but the definition everyone’s been quoting is Sun’s. She defined it, in her newsletter, as a worldview that treats AI as an elite political project: something tech billionaires are pushing onto an unwilling public in the service of layoffs and surveillance, with the question of whether the tools actually work treated as beside the point. It’s a sharp definition, and it spread fast. By early May, “A.I. Populism Is Here” was a New York Times headline. And then, the way these things go, the term became an argument about the term: Merchant pushed back on it, and pointed to the writer David Karpf doing the same, on the grounds that it lumps very different people together and makes their objection sound more paranoid than it is.
Underneath all of that, the question had changed. It had been whether the students were right. It became what to call them.
The students had made an argument. The commentary turned the argument into a category, and then turned the category into a debate about the category: whether “AI populism” is too broad a net, whether it condescends to the people it describes. None of that is stupid, and some of it I’d defend. But it is, unmistakably, the sound of an event being processed into a topic. A boo is a boo. “AI populism” is a panel discussion. (In fairness, this essay is one too.)
A boo with no teeth
So, the question. What can a boo actually do?
On the evidence here, not much. Nobody had to suppress it. Nobody cut a microphone or stopped a single graduate from booing. The boos were loud, they were clear, they went viral, and none of that mattered, because loud and clear and viral isn’t the same as having leverage. Send a boo through all three rooms and what comes out the other end is everything except a reason for anyone to do anything differently.
None of it took malice. Nobody in any of those rooms set out to neutralize a class of angry graduates; they just did the jobs they already had, and the jobs, run end to end, do the neutralizing on their own. Schmidt had told the graduates that “when someone offers you a seat on the rocket ship, you do not ask which seat, you just get on.” They booed him through most of the speech. He finished it, and the evening cost him nothing. Caulfield stood in front of thousands of people rejecting her and decided, out loud, that what she was hearing was “passion.”
The industry knows the difference between a boo and a threat. In Sun’s reporting, the Palantir CEO Alex Karp warned a panel that “the country could blow up politically.” He wasn’t talking about graduates booing. In April, someone tried to firebomb Sam Altman’s home; someone else is accused of shooting at an Indianapolis official who had voted for a data center. Those are crimes, and they are also the version of the backlash the industry takes seriously. A boo is the other version, the civil one, the kind that fits inside a thirty-second clip. It is also the most forceful thing a stadium of twenty-two-year-olds can actually do.
So the students were right, and they were heard. Unfortunately, it probably won’t matter.
If you made it this far, you've earned a little something. During the same stretch of graduations, Glendale Community College in Arizona let an AI system read its graduates' names at commencement. It skipped dozens of them. I went through and edited together a little video of the mishap. Enjoy!
Here’s my latest accepted comment at NYTimes re Dowd’s recent piece on AI:
The word "cannibalism" comes to mind. . . Nothing like ripping off almost the entire archives of humans with machines, then take all the profits from that enterprise and call it your own to sell and make available mostly behind paywalls.
If anything, any enterprise that is pretending to own all that data is a thief unless it shares all profits from such thievery with all humans on earth. A cut of the profits from our collective property. We created what it currently claims to harvest and own. Each and every one of us and the archives of ancestors.
Soylent Green should be the official Pantone color for 2027. To remind us of how much toxic capitalism has eaten us up.
The speakers’ comments would hit maybe a little differently if they weren’t coming from the very people who are making AI “inevitable” so they can make money from it.