Sam Altman sees an AI bubble forming

The AI valuations feel irrational. And that’s the diagnosis from someone who would know: OpenAI CEO Sam Altman. Altman, the face — and often the voice — of the AI boom, just warned that the AI frenzy is spilling past reason into bubble territory.

At a recent dinner with reporters in San Francisco — first reported by The Verge — he didn’t mince words. When Altman was asked whether investors are collectively overhyping the AI space, Altman said, “Yes.” He compared the current surge of excitement with the late-1990s dot-com boom, when investors piled into internet startups on the back of one undeniable fact: that the internet was a world-changing technology.

“When bubbles happen, smart people get overexcited about a kernel of truth,” Altman said. In his view, AI is that kernel: a genuinely transformative advance. The problem is the frenzy of cash chasing anything labeled “AI.” 

“If you look at most of the bubbles in history, like the tech bubble, there was a real thing. Tech was really important. The internet was a really big deal. People got overexcited. Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes.”

That duality — AI as both a once-in-a-generation breakthrough and a speculative mania — is Altman’s current reality. At the dinner, he criticized the sky-high valuations for startups that barely exist beyond a pitch deck, warning that “someone’s gonna get burned.” Even as he acknowledged bubble dynamics, Altman emphasized that OpenAI is pressing ahead at full speed. 

“You should expect OpenAI to spend trillions of dollars on data center construction in the not very distant future,” he said, betting that the infrastructure being built now will outlast whatever course correction comes later. Capacity constraints are already biting, he said: “We have to make these horrible trade-offs right now. We have better models, and we just can’t offer them because we don’t have the capacity.” 

It’s rare for a CEO at the center of a hype cycle to wave the caution flag, especially one leading the company that helped ignite it with ChatGPT. But Altman’s candor matches the contradictions of today’s AI economy — an industry that’s both delivering real products used by hundreds of millions and trading at valuations untethered from near-term revenue.

Altman isn’t the only one drawing parallels between the AI era and the dot-com era. Erik Gordon, a business professor at the University of Michigan and so-called “tech guru,” told Business Insider last week that the market reaction to CoreWeave’s 33% stock plunge in two days — a $24 billion loss — was an early crack. “More investors will suffer than suffered in the dot-com crash,” he said, warning that the fallout could be harsher this time. Apollo Global Management’s chief economist, Torsten Sløk, has gone further, warning AI stocks today are even more overvalued than those in the late ’90s. And earlier this year, Bridgewater founder Ray Dalio told the Financial Times that dangerously high valuations, coupled with elevated interest rates, could “prick the bubble,” making today’s market look eerily like 1999.

“There’s a major new technology that certainly will change the world and be successful,” Dalio said. “But some people are confusing that with the investments being successful.”

Wall Street strategists are sounding alarms, too. Bank of America’s Michael Hartnett flagged that the S&P 500’s price-to-book ratio has hit 5.3, topping dot-com levels, while forward P/Es hover around 1929 territory. Veteran investor Richard Bernstein has been urging clients to avoid chasing “AI FOMO” and instead rotate into dividend payers and defensive sectors such as utilities.

Not all the analysis is doom-laden, though. Economists have noted that AI spending is already showing up in GDP: Capital expenditures on chips and data centers are outpacing consumer spending as a driver of growth. That suggests that, even if valuations come down, the hardware and infrastructure being laid today may become the lasting legacy — similar to the fiber-optic networks that survived the dot-com bust. So it’s possible that AI isn’t just tech wizardry; it’s infrastructure fueling real economy momentum.

Altman’s conversation with The Verge and others also touched on OpenAI’s missteps with its latest release, ChatGPT-5. “I think we totally screwed up some things on the rollout,” he said. But he quickly pointed to the upside: API traffic doubled in 48 hours, GPU demand spiked, and ChatGPT “has been hitting a new high of users every day.” The scramble to roll back changes underscored, in his view, just how complex it is to update a product that’s used by hundreds of millions of people — overnight. 

But if GPT-5 was a stumble, Altman quickly pivoted to the bigger picture: ChatGPT’s relentless climb. Unsurprisingly, the CEO was bullish about ChatGPT’s trajectory, calling it one of the world’s dominant platforms already. “Pretty soon, billions of people a day will be talking to ChatGPT,” he said. “We’re the fifth biggest website in the world right now. I think we’re on the clear path to the third.” Surpassing Google, he admitted, would be far more difficult: “For ChatGPT to be bigger than Google, that’s really hard.”

Altman showed little hesitation in sketching out the next frontier, one that pushes far beyond search engines and into science fiction. Altman also confirmed OpenAI is exploring brain-computer interfaces to rival Elon Musk’s Neuralink. “I think neural interfaces are cool ideas to explore,” he said. “I would like to be able to think something and have ChatGPT respond to it.” He praised the company’s hardware collaboration with Jony Ive and hinted at a new computing paradigm: “You don’t get a new computing paradigm very often. There have been only two in the last 50 years. So just let yourself be happy and surprised. It really is worth the wait.”

But despite being the face of OpenAI, Altman said he doesn’t see himself as a natural long-term CEO. “I’m not a naturally well-suited person to be a public company CEO,” he said. “Can you imagine me on an earnings call?” Asked whether he expected to remain in the role in a few years, he said, “I mean, maybe an AI is [the CEO] in three years. That’s a long time.”

He may laugh off the prospect of an AI running OpenAI in three years, but he’s dead serious about the market’s froth. For now, the industry is floating on boundless optimism and bottomless cash. The trouble with balloons, of course, is that they don’t stay aloft forever.