Don't Accept the First Answer to a Vague Question
What it is
A habit correction: the first answer to a loose question is usually a generic starting point, not the best ChatGPT can do — refine it with context and specifics instead of taking it as final.
Why it works
A vague prompt gets a safe, middle-of-the-road answer because the model is hedging across everything you might have meant. That first reply is a draft, and its main value is showing you what to correct. People stop there and conclude ChatGPT is mediocre, when a single round of 'here's my actual situation, redo it' transforms the output.
When to use it
Any time the first answer feels generic, obvious, or not quite right — that's the signal to add specifics and iterate, not to give up.
When not to use it
Simple factual questions where the first answer is complete and correct, and there's nothing to refine.
Prompt
That's a good start but too generic for me. Here's my actual situation: <specifics, constraints, what you already tried>. Redo it for this — and if anything's still unclear, ask before answering.Example
A bland 'tips to improve your resume' becomes genuinely useful once you paste the resume and say 'I'm switching from teaching to UX' and ask again.
Common mistakes
- Taking a hedged first answer as ChatGPT's best and quitting.
- Complaining it's generic without giving the specifics that fix it.
- Re-asking the same vague question instead of adding context.