This clip went viral last week: Eric Schmidt, ex-Google CEO, getting booed at a university commencement speech.

Who would have thought that telling a bunch of new grads that AI is already taking a job they don't have yet would not be appreciated?

My favorite comment from the video says everything about how detached from reality someone like Schmidt (worth ~$60B) really is:

But this got me thinking about a real question: do people actually hate AI?

Specifically, would people outside of tech prefer to shut the whole thing down if they could?

Even the people who work most closely with AI have their complaints. Developers, for instance, now spend a good chunk of their day reviewing AI-generated code slop…

Gary Marcus, who has been a long-time “AI realist”, thinks this growing frustration is going to be one of the defining issues of the next US election cycle.

I'd recommend reading his latest piece, where he draws a pretty uncomfortable comparison between the current AI wave and the Vietnam War:

“The war moved forward for years despite evidence that it wasn’t working as planned. It was also ridiculously expensive, about a trillion in today’s dollars.”

Change a few words around and it sounds eerily familiar. Insane spending, negative returns, and models that aren't delivering on the exponential improvement that was promised.

The cheerleaders have moved the goalposts so many times it's becoming its own sport.

Talking to an AI enthusiast these days almost feels like talking to a crypto bro circa 2021.

Voice any concern and you either "just don't get it" or you're told this is the worst these models will ever be, so stop complaining. It's a closed loop designed to make skepticism feel stupid.

What's interesting is that more measured voices are starting to get traction. YouTuber Mo Bitar has been blowing up over the past few months for example.

And I think that gets to the heart of it. People don't hate what AI can actually do today.

Schmidt got booed not because the crowd was anti-technology, but because they were tired of being sold a story that conveniently benefits the person telling it.

That kind of sentiment has a way of getting political attention. When enough people are frustrated and vocal, governments tend to notice, and regulators who have been tiptoeing around the AI industry for years may find the political will to actually do something.

The concern then is that heavy-handed regulation, combined with cooling public enthusiasm and a growing list of unmet promises, creates the conditions for an AI winter.

It has happened before: periods where funding dried up, research slowed, and the whole space went quiet for years because the gap between the promise and the reality became impossible to ignore.