x
Humans vs LLM, memes as theorems — LessWrong