Should Garrabrant induction replace Bayesian epistemology? AI safety researchers debate
davidmanheim · x · 2026-11-04
davidmanheim follows up on Richard Ngo's thread about Garrabrant induction, saying he suggested years ago that logical induction deserved a formal epistemology paper so people would notice its significance.
More from AGI Musings
- Satire: In 2029 the Senate subpoenas an AI model while a Chinese model ships a better battery by lunch — r0ck3t23 · 2026-11-04
- Austrian survey: only 3-9% of students aged 11-20 never use AI; weaker students use it most — mitsuhiko · 2026-11-04
- Ben Goertzel: first superintelligence could 'conquer the planet'; AGI in 2-4 years — IndividualDot9604 · 2026-11-04
- DNS inventor Paul Mockapetris and Vint Cerf debate DNS as a discovery layer for AI agents — IvyTatiana88 · 2026-11-04
- AI literacy should be about emotional regulation, not prompting, says researcher — random_walker · 2026-11-04
- AI Lurches Towards RSI: GrokBot costs, model-editing-weights paper, Anthropic's S1 update — The AI Daily Brief · 2026-11-04