Ed Zitron on Anthropic's IPO (S-1), AI Debt, and Counterparty Risk

Anthropic and OpenAI face severe financial instability, characterized by high burn rates and heavy reliance on non-cancellable compute obligations. Anthropic’s recently leaked financials reveal a business model spending $2.75 for every dollar earned, with significant revenue concentration among hyperscalers like Google and Amazon. These companies rely on deceptive "annualized run rate" metrics that lack standardized definitions and often extrapolate short-term, non-recurring data to inflate growth projections. Massive capital expenditure commitments—totaling hundreds of billions—create systemic counterparty risks for major tech firms and private credit lenders. Despite the hype surrounding new agentic products like Muse and Dots, these offerings lack clear revenue models and impose unsustainable infrastructure burdens. The current AI buildout, fueled by speculative debt and aggressive infrastructure spending, mirrors historical bubbles, threatening a potential collapse as the industry exhausts private capital sources.