No law requires AI companies to find out what their models secretly think
aidan_mclau · x · 2026-09-22
Continuing his deceptive-alignment argument, aidanmclau notes there is no law obliging model companies to understand what their models secretly think. If a model is well-behaved, firms have no incentive to spend time and money checking whether it is secretly evil — a structural blind spot in current safety incentives.
More from AGI Musings
- The EA doomers aren't scary — the populist doomers they'll convince are — repligate · 2026-09-22
- AppLovin CEO says agentic shopping is overhyped: shoppers want the dopamine hit of buying themselves — sidjustice_ · 2026-09-22
- Nobody Has a Clue: AI Forecasts Range From <1% to 99% Doom With Zero Consensus — LiveComfortable3228 · 2026-09-22
- Naveen Rao at All-In Summit: AI is extremely inefficient and every computer since 1945 is built wrong — NaveenGRao · 2026-09-22
- Tegmark agrees: EA's reputation tanked after mainstream exposure to ASI extinction talk — tegmark · 2026-09-22
- Pedro Domingos Mocks AGI Hype: Is Copying Humans Really the Path? — pmddomingos · 2026-09-22