The article critiques apocalyptic AI statements from industry leaders as strategic rather than sincere. It traces the 'genre of utterance' from the May 2023 Center for AI Safety statement—signed by Sam Altman, Demis Hassabis, Dario Amodei, and others—to subsequent quantitative claims like Amodei's 25% doomsday figure. The author argues these statements, while personally sincere for figures like Evan Hubinger and Jacob Coxon, serve institutional purposes: they influence markets, regulators, and public discourse while crowding out nuanced discussion.

The piece highlights the hypocrisy in tactics, noting how Altman simultaneously urged US regulation while lobbying for EU exemptions. This 'doom archive' employs a stable grammar across 2023-2026, transforming existential risk into a rhetorical tool that legitimizes acceleration rather than genuine safety measures.