Zvi mocks 'show me one AI killing' argument against AI safety pauses

TheZvi · x · 2026-09-18

Responding to the claim that no one can point to a single instance of AI killing anyone and thus there's no reason to shut it down "for safety," AI safety commentator Zvi agreed ironically: it's true he cannot point to a single instance of AI killing everyone — and, he noted, some people suggest he should wait until he can. The exchange is a condensed version of the ongoing debate over precautionary reasoning in AI risk.

Original post →

More from AGI Musings

AGI Musings channel →