NYU professor warns AI bubble could be worse than dot-com crash

A finance professor at New York University, Aswath Damodaran, warns that if the AI industry bubble bursts, the impact could be worse than the dot-com crash in 2000. He says AI needs huge investments in data centers and other physical infrastructure, much of it funded by debt. If the market crashes, losses won't just hit shareholders but could spread to society.

Damodaran also doubts AI can achieve the scale benefits people expect. Unlike traditional software, every AI use costs more computing power, making its business model more like Spotify, which pays per play, than Netflix, which spreads costs over subscribers. If companies chase growth with thin profits, they might destroy value.

Even if the best AI stories come true, they could still cause serious problems. If AI replaces entire jobs, half of all white-collar workers could lose their jobs. Damodaran says if the big promises behind high AI valuations come true, the cost to society would be huge, and we need to start thinking about it now.