1 nota · última: 2 Sep 2025
The post argues that LLMs are not human-like minds but autocomplete systems that lack proper world models, so they keep making obvious errors in time, history, economics, and factual reasoning.