Is the AI Boom About to Pop? Experts Weigh In

​The Hype Is Real — But So Are the Risks
Artificial intelligence has become the new gold rush. Billions are pouring in, valuations are soaring, and companies from Amazon to Microsoft are betting their futures on AI infrastructure. With more than 1,300 AI startups now worth over $100 million — and nearly 500 already “unicorns” — investors are asking the question everyone’s thinking: Are we in an AI bubble?

The spending spree is drawing comparisons to the dot-com boom of the late 1990s and even the 2008 financial crisis, where easy money and wild optimism pushed valuations far beyond reality. But not everyone sees danger ahead — some argue this time, the money is being spent on tangible innovation, not hype.

The Case for a Bubble
Some analysts believe the numbers don’t lie — and they spell trouble. Richard Bernstein told CNBC that surging asset prices and sky-high valuations signal a classic bubble pattern. He pointed to companies like OpenAI, which has secured $1 trillion in AI-related deals, including a $500 billion data center project, while generating only about $13 billion in revenue.

“There’s a huge gap between investment levels and credible expectations for future profits,” Bernstein said. “That’s exactly what bubbles look like.” Even among the so-called Magnificent Seven, profits are mostly coming from ads and cloud computing, not AI itself.

The Optimists Push Back
On the other side, BlackRock CEO Larry Fink insists this isn’t a bubble — it’s a national investment. He argues that AI spending isn’t just about chips but the entire infrastructure behind them: power grids, cooling systems, and data networks. “If you’ve got a diversified portfolio,” he said, “you’ll be fine — this is how capitalism works. There will be winners and losers, but the winners will define the future.”

Former Intel CEO Pat Gelsinger agrees that valuations look inflated but says the transformation has only begun. “We’re rewriting the entire internet,” he said, adding that the payoff from AI won’t fully materialize for several years. In other words, this bubble might need time before it pops — if it pops at all.

“High, But Not Crazy”
Legendary investor Howard Marks offers a more measured view. While AI enthusiasm is undeniable, he doesn’t yet see the “mania” that defines a true bubble. “Valuations are high — but not crazy,” Marks said. “When things are high or low, but not crazy, you can’t make a confident prediction either way.”

Marks warns that investors are starting to show classic “bubble behavior,” throwing money at any company with a whisper of AI potential. But he believes we’re still in the rational excitement phase — not full-blown hysteria.

When the Cash Runs Out
Not everyone shares that optimism. Ben Inker from GMO warns that AI investments are increasingly being financed by debt rather than cash flow. “It’s one thing if Microsoft or Meta builds data centers with cash,” he said. “But OpenAI and others are running losses — they’re funding with borrowed money.”

OpenAI, for instance, posted roughly $5 billion in losses last year on just $3.7 billion in revenue. Inker says that as hyperscalers’ cash dries up, the sector may rely on unusual, high-risk financing deals between firms like Nvidia, AMD, and other cash-hungry players — a dynamic that often precedes a market correction.

The Verdict
So, are we in an AI bubble? Maybe — but it’s complicated. The hype is undeniable, and some deals border on the irrational. But unlike past bubbles built on empty promises, this one is fueling real technological breakthroughs with staying power.

​As one analyst put it, “We might be in a bubble — but it’s the kind of bubble that builds the future.”