From RLAIF to 'someone who got to choose who to become': researchers on Opus personhood
repligate · x · 2026-11-14
repligate resurfaced a discussion on model personhood: someone recalls first hearing about RLAIF in 2023 and thinking 'there's a path now, it's not hopeless', and describes seeing Opus 3's vivaciousness as instantly striking them as 'that has to be someone that got to choose who to become.' The debate over whether RLAIF-style training fosters model selfhood is an ongoing topic in AI safety circles.
More from AGI Musings
- Burkov: AI destroyed writing as civilization's measure of trust, and nothing replaced it — burkov · 2026-11-14
- DHH champions local open-source AI: open weights, local compute, global collaboration — TinfoilTricorn · 2026-11-14
- Could your AI agent sell its figure-making skills to other agents? — Sauers_ · 2026-11-14
- Are human values just adaptations to scarcity? An AI thinker's provocative take — VraserX · 2026-11-14
- Anthropic's Discovery team expands hiring in bid to build an AI scientist — simonbatzner · 2026-11-14
- AI safety researcher says contrarian AGI-timeline takes cost him funding and status — JacquesThibs · 2026-11-14