Dan Luu Audits Ed Zitron’s AI Calls

The sharp takeaway is that skepticism becomes unreliable when temporary limitations are presented as permanent ceilings. Zitron often identifies real economic risks, but his high-confidence conclusions repeatedly outrun the evidence. Luu marks predictions about peak AI, exhausted training data, OpenAI’s stalled growth, DeepSeek commoditization, and Cursor’s demise as wrong or unresolved. Meta, Alphabet, and Microsoft continued growing, while OpenAI exceeded major revenue expectations and Cursor eventually reached a multibillion-dollar exit. The post also challenges Zitron’s methodology, including selective metrics and spreadsheet errors that undermine otherwise forceful financial arguments. For developers, Zitron is more useful as a source of questions about AI economics than as a forecast of model capability, adoption, or industry timing.