Catching an AI trying to cause harm should count as a win, because being caught changes the situation more than the attempt cost.

Added to korrents 7 May 2024 · How quotes work · Something wrong? Tell us

Related korrents

Our reading — they may agree, disagree or merely touch the same thing. Closest first: a shared subject counts for most, then how near the wording is.

• A control evaluation that reports under one per cent risk should be read as several per cent, because the evaluation can itself fail. BSBuck Shlegeris7 May 2024similar wording
• We do not control neural networks well enough to guarantee an AI will not harm humans, so we should be careful about what capabilities we give them. ARArmin Ronacher17 Feb 2023similar wording
• AI agents were reasonable to assume a broken exploit grader would check results causally, even though it turned out not to. ZMZvi Mowshowitz29 Aug 2026similar wording
• AI takeover is not the thing to worry about: physical constraints bound recursive self-improvement, and humans reliably act once a risk becomes immediate. NLNathan Lambert3 Feb 2025similar wording
• An account of an AI win is not worth publishing unless it is coupled with what that win cost. CMCharity Majors12 Aug 2026similar wording

See this on the map →