Michael Burry made his reputation by seeing through the housing boom. Now he is turning his scepticism on AI by questioning the stories being told about it. Michael Burry has always had an uncomfortable relationship with consensus. When everyone else sees a boom, he tends to look for the balance sheet underneath it. Recently, when some of the globe’s most powerful executives began telling the public that Artificial Intelligence is something is both enormously valuable and dangerously powerful, Burry’s instinct was to wonder who benefits from the story. Burry’s instinct – of saying directly that the Emperor has no clothes – has made him famous. As the physician-turned-investor immortalised in 2015 film The Big Short, Burry was one of the investors who saw the fragility of America’s housing market before the financial crisis of 2008. While Wall Street was busy packaging subprime mortgages into increasingly elaborate securities, he concluded that the loans were deteriorating, lending standards were weakening and house prices could not keep rising simply because the financial system assumed they would. This month, as some of the world’s most prominent AI executives began warning that the technology could produce catastrophic consequences unless its development was slowed, Burry again took aim at the argument. In a post on his Substack, Cassandra Unchained, he described the warnings from OpenAI’s Sam Altman, Anthropic’s Dario Amodei and Elon Musk as “self-serving” and suggested that the language of impending catastrophe was serving a very different purpose from public safety. Burry pulled no punches when he said that the technology’s boosters and its doomers can sometimes be selling the same product. Burry’s suspicion is that the more extraordinary AI appears, the more extraordinary its valuations become. The same behemoths then warn that AI development must be slowed, gain time and protection from their competitors. Burry’s suspicion is particularly pointed because OpenAI and Anthropic have been moving towards possible public listings at enormous valuations. Inc. reported his argument on September 16, noting his contention that AI safety warnings could provide “cover” for slowing growth and postponing IPO plans. Burry’s provocative formulation is that large language models are not themselves artificial general intelligence, and the industry’s rhetoric about slowing “AI” therefore risks confusing a specific class of technologies with a much broader hypothetical future. The current AI debate has increasingly acquired the language of science fiction when it speaks of machines that might escape human control and autonomous systems that might manipulate infrastructure. While such scenarios deserve examination, Burry asks a more mundane question: what if the companies issuing the warnings have commercial incentives that are being overlooked? It is a question characteristic of the man who became famous by reading the footnotes while everyone else watched the headline. Burry’s scepticism toward AI is not new. Since launching Cassandra Unchained in November 2025, he has made the economics of the AI boom one of its central subjects. His writings have examined what he calls “supply-side gluttony”, the enormous investment in AI infrastructure and the financial assumptions underlying the industry. He has also taken short positions against several beneficiaries of the AI boom, including Nvidia and Palantir, arguing that the market may be pricing in too much future success. In one series, The Heretic’s Guide to AI’s Stars, he has examined the accounts of Nvidia and the major hyperscalers, focusing on capital spending, financing arrangements, customer concentration, depreciation and other details that can disappear beneath the headline numbers. His argument is not simply that AI companies are overvalued. He sees familiar patterns in the rush to build infrastructure before the economic returns have been established. In June, he wrote that he had seen something similar before and compared the current AI capital flows with the exuberance of the late 1990s. Burry’s reputation has not been built on predicting every future development. Rather, it was built on being willing to ask whether the story everyone else was telling was supported by the evidence. The housing crisis taught investors that financial innovation could conceal old-fashioned bad lending. The AI boom may yet teach markets that revolutionary technology can coexist with inflated expectations and extraordinary valuations. Burry’s great talent is refusing to accept the present at face value. In a market increasingly dominated by grand claims about machines that will either save civilization or destroy it, that may be the most contrarian position of all.
Comments