One of the stranger things you learn when experimenting and talking to people about ChatGPT is how often the AI faces vague questions hiding hopes for very specific answers.
Intending to ask for help with something practical and ending up with a pile of loosely connected frustrations dumped into the prompt box is surprisingly common. The request exists somewhere in there. Unfortunately, it is often buried underneath opaque references, poor overexplaining, and other vagaries of language. But of course, ChatGPT responds only to the literal request instead of the actual problem sitting underneath it.
