Nathaniel Whittemore reviews arguments about an AI investment bubble, distinguishing a fall in stock prices from a wider credit crisis. He compares warnings from analysts with more optimistic views and notes the financial interests of people making the claims.
He considers data center financingAI compute infrastructure is the hardware, facilities, networks, power, cooling, storage, and software used to train and run AI models. and reports of GPUs being stored before deployment. He treats power and construction delays as a plausible explanation while identifying changes in financing or demand that would make him more concerned.
Whittemore discusses exposure through insurers and pension funds, customer concentration and upcoming public financial disclosures. He presents his assessment as uncertain and does not claim that a crash or systemic crisis is inevitable.
Watch on YouTube
