Odd LotsCohost Joe Weisenthal Has Predictions About How the AI Bubble Will Burst (www.wired.com)

Bloomberg Odd Lots cohost Joe Weisenthal warns that talk of an imminent “AI bubble” should be understood alongside broader macro dynamics: massive capital flowing into data centers and AI infrastructure is creating second‑order distortions across the economy that could amplify any market correction. He argues the risk isn’t just a valuation pullback but that concentrated demand for physical infrastructure—servers, cooling, electrical gear and gas turbines—is crowding out other industries, driving persistent shortages and elevated prices. The phenomenon is visible in ISM manufacturing surveys, where electrical equipment has been reported as tight for roughly five years and order books for critical parts extend toward 2030. For the AI/ML community this matters practically and strategically. Deployment depends on scarce hardware and power infrastructure; supply‑chain bottlenecks and price signals (rather than planned allocation) favor big data‑center builders over smaller projects, raising costs and fragility if AI spending slows. Underinvestment in other sectors (like housing or local facilities) compounds the problem, and geopolitical moves such as tariffs further complicate material flows. In short: even if models and algorithms advance, the pace and resilience of real‑world AI scaleup will hinge on constrained physical capital, hardware markets, and macroeconomic feedbacks that could magnify a downturn.