Comment by capl

That’s funny cause I feel the opposite: LLMs can automate, in a sloppy fashion, building the first trivial draft. But what remains is still thinking hard about the non trivial parts.