Comment by numpad0
IME, it's better to just delete erroneous responses and fix prompts until it works.
They are much better at fractally subdividing and interpreting inputs like a believer of a religion, than at deconstructing and iteratively improving things like an engineert. It's waste of token count trying to have such discussions with an LLM.