a day ago
- NVIDIA overstated Blackwell GPU efficiency gains, and AI nonetheless fails to make startups profitable.
- Hyperscalers have dramatically increased debt and capital expenditure for AI, with no disclosed AI revenue and growing cash-flow negativity.
- AI data center spending creates a vicious cycle: more demand raises hardware prices, requiring more debt, which becomes more expensive.
- Bond spreads for AI-related companies like CoreWeave indicate market anxiety about repayment ability.
- Circular financing deals, such as NVIDIA’s proposed guarantee for SoftBank’s SB Energy, rely on unsustainable financial structures.
- Outside of leading AI labs, demand for gigawatt-scale compute is minimal, and most AI products fail to generate meaningful revenue.
- Hyperscalers cannot stop AI capex without revealing poor returns, but continued spending deepens their financial hole.
- AI's practical benefits are limited; it functions more as a symbolic tool for executives than a genuine productivity driver.
- The AI industry is sustained by hype and media complicity, not actual economic value, risking a painful correction.