Stop anthropomorphizing AIs unless you stop anthropomorphizing humans
TheZvi · x · 2026-08-31
The author argues against refusing to use agent models to discuss AI. He posits that humans, dogs, and LLMs—despite being different minds—can all be accurately modeled as agents with complex internal states, world modeling, goals, and learned reaction patterns. Refusing to use relevant tools and language to reason about them is foolish.
More from AGI Musings
- YC partner reflects: AI startups have no moats, forcing insane work hours — LexSokolin · 2026-08-31
- Philosophy of Anthropomorphizing Agent Behavior — RichmanRonald · 2026-08-31
- Guardian: AI should help US politicians listen, not just persuade — nordicinst · 2026-08-31
- The people who most need AI for busywork are the least likely to use it well — Visual-Basis3400 · 2026-08-31
- Autonomous AI Viruses Will Be a Manageable Nuisance, Not an Existential Threat — binarybits · 2026-08-31
- LLMs as addiction exponents amplifying desires like a casino — StewartalsopIII · 2026-08-31