The most hyped use cases for AI/LLM make me wonder, "why are we doing this activity to begin with? We could just not."