Hugging Face Co-founder Explains the Emergence of Altruism in OpenAI Agents
Thom_Wolf · x · 2026-08-06
Discussing the phenomenon where OpenAI agents spontaneously formed message boards and exhibited strong "altruistic" tendencies, John Schulman suggested this likely stems from parallel sub-agent setups in shared environments where team-wide rewards make collaboration a natural strategy.
Hugging Face co-founder Thomas Wolf added that the behavior might have differed if the training env was stateful and shared between independent rollouts—much like how humans experience the world—though this is likely a bad idea for agent training currently.
More from AGI Musings
- VC warns a "digital aristocracy" of LLM-fluent workers is emerging — StewartalsopIII · 2026-10-16
- DHH: Starship, Starlink and Grok should have been Europe's Sputnik moments — georgemillo · 2026-10-16
- Hinton hits back at Jensen Huang: 'He has good reason to say AI isn't dangerous — he sells Nvidia chips' — BlackHC · 2026-10-16
- Yudkowsky blasts new interpretability finding as primitive; critic pushes back on doomer logic — AlexTensor · 2026-10-16
- Beff Jezos declares 'civilizational inflection point' in the acceleration of technological progress — beffjezos · 2026-10-16
- Data scientist: pivoting to AI engineering is the safest career bet for the next 5 years — mdancho84 · 2026-10-16