Prompting is not the point

Prompt-writing gets treated as the headline skill of working with AI, and it is the easiest one to teach, which is probably why. It is also the least interesting one to assess, because a better prompt mostly just gets you a better draft — it does not tell you whether the person asking would recognise a bad one.

The skill worth testing sits earlier: knowing roughly what a good answer would look like before you ask the question. That expectation is what lets someone spot a plausible-sounding wrong answer, push back on a shallow one, or recognize when the model has quietly changed the question it is answering.

That is domain knowledge and judgement, not prompt syntax — and it is why AIQQA's qualifications spend far more time on the thinking around the tool than on the tool itself.