Nobel economics laureate warns chip sell-off may point to AI ‘quasi-bubble’ starting to burst

The plunge in semiconductor stocks across the United States and Korea this week could be an omen of "a quasi-bubble quasi-bursting," according to Nobel Prize laureate economist Paul Krugman.

Krugman, a professor at the City University of New York, made the assessment in a video commentary posted Wednesday, after a chip-led sell-off swept markets on both sides of the Pacific earlier this week.

The Philadelphia Semiconductor Index fell almost 8 percent on Tuesday, the Kospi dropped about 10 percent and the Nasdaq lost 2.2 percent. Samsung Electronics and SK hynix, long among the biggest beneficiaries of the global AI buildout and the heaviest weights on the Kospi, were caught in the rout.

The mood around AI has turned sharply, Krugman said.

"The tone, the rhetoric surrounding use of AI, and hence the demand for compute, has really shifted quite a lot just very recently," he said.

He pointed to a spate of recent studies suggesting AI's real economic payoff falls well short of its output.

"AI models allow people to churn out a lot more stuff, but the actual payoff to that stuff is much, much smaller than the volume of stuff that they're churning out, most obviously lines of code, but just in general," he said.  "AI lets you do much more, but how productive that is in terms of the ultimate goals of a business, let alone economic growth and quality of life, is much more doubtful."

Companies have reversed course on how hard they push the technology, according to the researcher. Until recently, he said, businesses were "more or less whipping their workers into using AI," grading them on how much they used it.

But with computing power growing scarce and chip prices soaring, he said, firms have made "a sudden U-turn," now urging staff to cut back on usage to rein in demand for computing power.

Krugman cast the boom as more fragile than an ordinary market mania.

"It's a kind of a bubble, but not in the normal sort of asset price form," he said. "It's more of a kind of fad, almost a social delusion. And that, it seems likely, certainly got ahead of itself."

Much of it, he argued, has been driven by pressure rather than real market demand. "Fear of missing out, not by the individual investor, but by the corporate bureaucracy," Krugman said. "And then pressure from the financial markets, saying, you know, your company better be on the cutting edge of AI or else."

Even so, he warned against reading too much into a single week. The Philadelphia Semiconductor Index had climbed 157 percent over the past year before the drop, he noted, and the Kospi was up 172 percent over the same period, even after its 10 percent fall. "We're not talking about a catastrophe," he said. "But okay, it's a break in the trend."

Krugman also warned that markets often mislead.

"The famous old line by my teacher and colleague, Paul Samuelson, was that the stock market had predicted nine of the last five recessions," he said. "There's many more than that now. In fact, just over the course of the past year and a half, we've had two major stock market declines that turned out to be false alarms."

Both proved milder than feared, he noted. The Donald Trump administration's 2025 tariffs "did not cause an economic catastrophe and stocks recovered the losses that they experienced then," Krugman said. The war with Iran was a debacle in its own right, he added, but "the consequences for short-run macroeconomics were more modest than many people, myself included, expected."

The sell-off also landed amid an unmistakable shift in tone, Krugman said. He pointed to a recent interview with Microsoft chief executive Satya Nadella, noting that the company is a consumer of AI rather than a producer.

"Nadella was pretty scathing about saying [...] we can't give all of this power and all this money to the big AI companies, and we should be using cheaper models," Krugman said. Nadella also hinted that Microsoft might turn to DeepSeek, the Chinese model that is less comprehensive than its U.S. rivals but far cheaper and runs on much less computing power, according to Krugman.

What the markets may be signaling, he said, is not disappointment in AI itself but a recognition that today's compute-hungry approach is not essential. 

"Maybe you can still get whatever the big productivity benefits are and still possibly the big labor-displacing effects without quite so much compute," he said. "I think we need to be saying that this is what a quasi-bubble quasi-bursting might look like."

BY PARK JONG-SUH [[email protected]]