The AI industry faces a significant structural crisis as the transition from bundled subscriptions to metered, token-based pricing exposes the true, high costs of AI operations. Companies are building massive, debt-funded data centers based on the assumption of perpetual growth, creating a dangerous duration mismatch between long-term financial obligations and a deflating commodity. Despite the hype, AI productivity gains remain largely unmeasurable and often result in the production of low-value content rather than meaningful economic output. The current reliance on faith-based arguments—assuming AI will inevitably follow the path of previous technological revolutions—ignores the unique fragility of AI infrastructure and the lack of tangible return on investment. As hyperscalers become the largest issuers of investment-grade debt, the industry risks creating stranded assets that will collapse when the speculative bubble finally bursts.
Sign in to continue reading, translating and more.
Open full episode in Podwise
