Michael Burry, the investor best known for betting against the US housing market ahead of the 2008 financial crisis, has sharpened his warning about artificial intelligence stocks, saying the bubble he has been describing “may burst” sooner than he previously believed.
The comment, reported by CNBC, marks an escalation in tone rather than a change in position. Burry has spent much of the past few years arguing that enthusiasm for AI has outrun the underlying economics, pushing valuations for chipmakers, cloud providers and model developers to levels that assume years of flawless execution and uninterrupted demand. What has shifted, by his account, is the timing: the reckoning he expected at some indefinite point in the future now looks nearer.
Why the warning lands differently
Burry is not the only skeptic in the market, but he carries a particular kind of authority. His housing bet, immortalised in Michael Lewis’s The Big Short and the film adaptation, made him a shorthand for the contrarian who sees the crack before anyone else. That reputation cuts both ways. Burry has also issued bearish calls that did not pan out, or that arrived far too early to be profitable, and he has at times deleted his public commentary shortly after posting it. Investors have learned to treat his warnings as provocations worth thinking about rather than trade instructions.
Still, the substance of the AI skepticism is familiar and widely debated. Critics point to the enormous capital expenditure required to build and run data centres, the difficulty of translating impressive model capabilities into durable profit, and the circularity of deals in which suppliers, customers and investors in the AI supply chain are often the same handful of companies. Bulls counter that AI is already generating real revenue, that productivity gains are compounding, and that the biggest spenders are among the most cash-rich businesses in history.
The bubble question
Calling something a bubble is easier than calling the top. Even market participants who agree that valuations are stretched disagree violently about when gravity reasserts itself, and history is full of expensive lessons for those who shorted a mania a year or two early. The dot-com comparison, invoked constantly in AI discussions, works in both directions: the internet did transform the economy, and a great many internet stocks still lost most of their value first.
Burry’s revised timeline, if it proves accurate, would matter well beyond technology portfolios. AI-linked names have accounted for an outsized share of index gains in recent years, meaning a sharp repricing would drag on passive investors who never made an explicit bet on the theme. Capital spending tied to data centre construction has also become macroeconomically significant, touching energy, construction and industrial supply chains.
What to watch
For investors trying to assess the claim on its merits rather than on the messenger, the signals to monitor are relatively mundane: whether hyperscaler capital expenditure guidance holds or is trimmed, whether AI revenue growth at software companies converts into margin expansion, whether financing for new compute capacity stays cheap, and whether the concentration of market gains narrows further or broadens out.
Burry has made a prediction about direction and, now, a looser one about timing. The market, as ever, will settle the argument on its own schedule. Read More
Leave a Reply