Ben Todd calls out OpenAI's circular logic: automating AI research to defend against the superhacking AI it creates

ben_j_todd · x · 2026-09-23

80,000 Hours founder Ben Todd mocked OpenAI's safety rationale: OpenAI argues AI research must be automated to build AI capable of defending against a superhacking AI. Todd counters that the urgent threat wouldn't exist if OpenAI didn't automate research and build it in the first place — highlighting the self-justifying loop where labs cite new risks as justification for accelerating the very work that creates them.

Original post →

More from AGI Musings

AGI Musings channel →