A critical roundup of signals suggesting AI productivity gains are overhyped. Key points: a DX longitudinal study found only ~10% productivity improvement from AI tools (not the promised 10x); notable AI boosters like Gergely Orosz and Steve Yegge are publicly acknowledging downsides; GenAI companies have unsustainable unit economics where every user interaction loses money; insurers are lobbying to exclude AI-related liability from business policies; and Windsurf's drastic pricing changes illustrate the financial pressure on AI tooling companies. The author argues solid software engineering practices matter more than AI tools, and that the industry's financial reporting is misleading investors.