Who evaluates the independent evaluators? Philosopher flags oversight gap in AI safety reviews
connoraxiotes · x · 2026-09-13
Philosopher Jeff Sebo raises a recursive problem for AI safety: as frontier labs rely on "independent evaluators," who audits the auditors?
Even without outright collusion, institutional relationships — funding and partnership ties — can compromise review integrity. His proposed remedy is meaningful democratic oversight: public accountability for how evaluators are selected and supervised, rather than relying on industry self-regulation alone.
More from AGI Musings
- AI Folks Mull Rogue Agents Facing the Same Costly-Survival Economics as Humans — voooooogel · 2026-09-13
- Should antitrust be relaxed for frontier AI labs heading toward a cartel? — jessi_cata · 2026-09-13
- NYT: Doomsday conversations snowball inside Anthropic, OpenAI, Meta and Google — pstAsiatech · 2026-09-13
- Pacing frontier models is being debated by a public that has never seen one — theteknosaur · 2026-09-13
- Gordic Aleksa Dissects Kokotajlo's 'AI 2040' Plan A: A Chain of Low-Probability Assumptions — gordic_aleksa · 2026-09-13
- Debate on AI safety: fight bad AI with wiser AI, or curb misuse first? — DanielSMatthews · 2026-09-13