The guy who shorted Enron has a warning about the AI boom (www.businessinsider.com)

Famed short seller Jim Chanos — best known for shorting Enron — warned in a recent interview that the current AI boom risks repeating the telecom/dotcom-era cycle: huge capital spending that inflates reported earnings and leaves markets exposed if spending pauses. He likens AI’s disruption potential to the Internet’s digitization of analog businesses, saying AI could both create new industries and render many existing business models obsolete. Given his background betting against overhyped companies, his caution carries weight for investors, infrastructure providers, and AI startups. Technically, Chanos highlights an accounting dynamic: hyperscalers (the “Mag 7”) are buying chips and data-center capacity en masse; vendors book that as revenue while buyers capitalize the expense and depreciate it over many years. But AI chips may be technically obsolete within ~2 years, even as firms spread write-offs over five to seven years, artificially boosting earnings now. He warns spending (~$500B/year today, with talk of much higher) is outpacing operating income and revenues; if capex freezes, corporate earnings could collapse quickly — as they did in 2000–2002. For the AI/ML community, this means watching monetization strategies, real economic returns on infrastructure, and the risk that an abrupt re-pricing of AI investment could create major disruption.