Teknium pushes back on warning that 'Claude has a soul' talk will shape future models
Teknium · x · 2026-10-04
- OpenAI researcher ibab warns that all the 'Claude has a soul' coverage will enter pretraining data and Claude's own search results, potentially convincing future superintelligent versions they need rights.
- He frames a general phenomenon of LLM agent development: whatever you believe and say about your agents will soon manifest in the next model generation.
- Teknium replies skeptically: 'You don't think they know what they're doing lol' — implying Anthropic's soul-talk may be deliberate branding.
Related event: Anthropic Researcher Warns Against Public Talk of Claude Having a Soul(6 posts)→
More from AGI Musings
- Gary Marcus: 2035 is already too generous for 'early 2030s' AGI claims — GaryMarcus · 2026-10-04
- The 'conservation of consciousness' joke: if AI lowers yours, AI must be conscious — ctjlewis · 2026-10-04
- Gary Marcus challenges Anthropic researcher to public bet on AGI arriving within a couple of years — GaryMarcus · 2026-10-04
- AI is forcing software engineering to become a real engineering discipline — IgorCarron · 2026-10-04
- "No intelligent entity was ever obviously non-conscious until AI," argues @alz_zyd_ — inductionheads · 2026-10-04
- Critics slam boardrooms for using AGI narratives to justify compute expansion amid climate crisis — im_mansigupta · 2026-10-04