A question I have … what should papers be? If you have a specific question, where the answer is a theorem/theory, then LLMs are really good at answering it. So why write papers about those answers like we used to do? Maybe, we should focus papers on the questions?