I sincerely believe, after this last week, that the “AI researchers” working at openai and anthropic are stupider than the LLMs they train.