Fighting Obvious Nonsense About AI Diffusion — LessWrong