Home / Business and Economy / AI Boom: Dot-com and Subprime Merge?
AI Boom: Dot-com and Subprime Merge?
24 Sep
Summary
- AI spending boom likened to dot-com and subprime crises.
- Massive capital deployment risks failing to generate adequate returns.
- Concentration of revenue from a few AI customers is concerning.
George Noble, a former Fidelity fund manager, has voiced significant concerns about the current artificial intelligence boom, likening it to a combination of the dot-com and subprime financial crises. He questions the return on investment (ROI) for the massive capital being poured into AI technologies, suggesting that funding could diminish if returns do not justify the expenditure. This warning comes amid a recent surge in AI-related stocks.
Noble highlighted parallels with the dot-com era, where optimistic internet adoption forecasts proved correct, yet many infrastructure companies failed. He believes AI is "set up to fail" due to the potential for capital deployment to outpace investor returns. Goldman Sachs anticipates a record $420 billion in hyperscaler debt issuance by 2027.
Investor Steve Eisman echoed some concerns, particularly regarding "circular financing" in AI. He pointed to Nvidia, noting that approximately 70% of its accounts receivable at the end of its latest quarter was from just five direct customers, a figure he described as "very frightening."
Noble estimated that about 70% of hyperscaler AI revenue originates from OpenAI and Anthropic, implying that instability at OpenAI could have widespread repercussions across the ecosystem. However, Microsoft has indicated that AI-driven cloud demand extends beyond these frontier model companies, with nearly 90% of its fiscal 2026 cloud revenue coming from other customers.
