dioscuri clarifies: LLMs are anthropomimetic by default, humanlike reports aren't optional

dioscuri · x · 2026-09-17

Reacting to debate over LLM inner lives, dioscuri clarifies his position: LLMs are anthropomimetic by default — base models act like humans and report inner lives, but that doesn't mean taking reports at face value; the aim should be accurate self-reports. He adds he's open to psychological complexity in LLMs and that anthropomimesis isn't a debunking explanation: current methods produce humanlike behavior and reports by default, not as an optional extra.

Related event: Researchers Argue LLMs Are Anthropomimetic by Default(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →