> It is both shorter and much easier to understand than anything the LLMs generated. I think this makes sense. LLMs are trained on average code on average. Thi

> It is both shorter and much easier to understand than anything the LLMs generated. I think this makes sense. LLMs are trained on average code on average. This means they produce average code, on average.