The Bubble Question Gets a Surprising Answer

Are we in an AI bubble? According to investors at the Fortune Brainstorm AI conference, the answer is probably yes—and they're surprisingly okay with that.

Steve Jang of Kindred Ventures didn't mince words: "I think it is a bubble, but bubbles are good for innovation." His reasoning? Market excitement is what pulls top engineers out of comfortable positions and gets ambitious projects funded. Without some irrational exuberance, the really audacious ideas never get off the ground.

Jang pointed to engineers leaving cushy jobs at Alphabet Inc. (GOOGL) (GOOG), Meta Platforms Inc. (META), and Uber Technologies, Inc. (UBER) to launch their own startups. That talent migration, he argued, is a "good signal" for the industry—a sign that the best and brightest think there's real opportunity beyond the established giants.

Cathy Gao of Sapphire Ventures acknowledged that some valuations have clearly outpaced fundamentals. But she made an interesting counterpoint: AI growth curves "far outstrip the growth curves of companies we've ever seen before." When you're dealing with something genuinely unprecedented, traditional valuation frameworks start looking inadequate. The true potential remains unclear, making broad predictions difficult.

Where Smart Money Is Placing Bets

So if it's a bubble, where should investors focus? Jang said Kindred Ventures concentrates on AI infrastructure—the chips, GPUs, and cloud systems that underpin everything else. The margins there remain strong, and someone has to build the picks and shovels for this gold rush.

Gao emphasized enterprise applications, but with a crucial caveat. Simple "AI for X" solutions are vulnerable to competition and commoditization. The real opportunity lies in embedding AI deeply into complex workflows where it creates sustainable competitive advantages that are hard to replicate.

Robotics Reality Check

The conversation turned cautionary when discussing robotics startups. Jang warned that many are building on early-stage models that will inevitably improve—leaving those startups scrambling. "A whole bunch of robotics startups … are going to have a lot of heartbreak when the models improve," he predicted.

Gao added that even as AI models advance, enterprise adoption will face hurdles around trust and visibility. Companies need to understand what these systems are doing before betting their operations on them.

The Skeptics Weigh In

Not everyone shares the optimism. Last week, investor Michael Burry and Google DeepMind CEO Demis Hassabis raised red flags about overvaluation risks in the AI sector.

Burry said the AI bubble's timing was unpredictable and criticized Nvidia Corp. (NVDA) for fueling hype. He highlighted a particular concern: companies are overspending on data centers without corresponding real demand to justify the investment.

Hassabis noted the absurdity of some AI startups raising tens of billions of dollars before even launching products. He contrasted these newcomers with established tech giants like Google and offered a nuanced view: AI is "overhyped in the short term" but underappreciated in the medium to long term. It's a classic pattern where hype inflates valuations rapidly, but the real transformative impact takes years to materialize.

So yes, we're probably in a bubble. But if that bubble is what it takes to fund the next wave of innovation and convince the world's best engineers to take risks, maybe that's not the worst thing. The question isn't whether valuations are rational—it's whether the innovation they're funding will eventually justify the hype.