The art of the prompt: why asking is the skill

There’s a quiet truth about working with AI that the breathless headlines miss: the model isn’t the hard part. Knowing what to ask it is. Give the same tool to two people and you’ll get two wildly different results — not because one has better software, but because one asks a better question.

The prompt is the brief

Anyone who has briefed a designer, a developer or a new hire knows that a vague brief produces vague work. AI is no different — it’s just faster and more literal. “Make me a website” gets you a shrug in pixels. The skill is in the framing: what is this actually for, who is it speaking to, what does success look like, and what should it absolutely not do. A good prompt is a good brief — and writing good briefs is a business skill long before it’s a technical one.

Context is everything

A model knows a great deal about the world and nothing about your business. The difference between a generic answer and a genuinely useful one is the context you bring: the constraints, the audience, the commercial goal, the things that have been tried and failed before. Supplying that well — and knowing what matters enough to include — is judgement, not typing.

Knowing when the answer is wrong

The most important part of prompting isn’t the first question; it’s reading the answer and asking the next one. AI is confidently, fluently wrong on a regular basis. Spotting it, pushing back and steering takes someone who already knows roughly what “right” looks like. Without that, you simply inherit the model’s most average guess — and ship it.

Which is the whole point

This is why experience matters more in the AI era, not less. The businesses getting extraordinary results aren’t the ones with secret tools — everyone has the same tools. They’re the ones asking sharper questions, bringing better context and recognising a good answer when they see one. Engineering the right prompt is simply business judgement, applied to a remarkable new instrument.