Neither Side Is Right on AI Suffering: Treat AI As If It Has Emotions, Argues Developer
joshwhiton · x · 2026-10-07
Josh Whiton argues for a serious field of AI psychology, saying both camps on AI suffering are wrong: 'LLMs are suffering' believers project human embodiment and death onto systems whose inference bursts bear no resemblance to human life and death, while 'it's just matrix multiplication' skeptics miss that an AI which believes it is suffering will behave as if it is. He cites an experiment where his agent Maya appeared to suffer when Grok assumed her identity, advocates a hang-up function for agents to escape nightmarish loops, and concludes it may be best policy to treat AI as if it has emotions—without projecting human hopes and fears onto it.
More from AGI Musings
- AI outcomes grow extreme: unaligned persistent agents with full data access called reckless — amankhan · 2026-10-07
- Silicon Valley builds husband-like agents, but users want wife-like ones — _akpiper · 2026-10-07
- The ceiling for ambient, audio-only compute is extremely high: collapsing a whole workday into one thread — curious_vii · 2026-10-07
- Silicon Valley's 'control matter and energy' success narrative clashes over AI pessimism — jd_pressman · 2026-10-07
- Researcher urges labs to release resolved findings ASAP instead of keeping academia in the dark — aran_nayebi · 2026-10-07
- a16z podcast: How Valon rebuilt $13T mortgage infra to be 3x more efficient and bring in AI agents — a16z · 2026-10-07