Prompt engineering
Also called prompting
The skill of writing requests that get an AI to do what you actually want: the context, the task, the format and the reasons.
The model can only work with what's on the page. It doesn't know your audience, your deadline or what you'll do with the answer unless you say so, so it fills the gaps with the most average guess.
Good prompting closes those gaps. Say who it's for and why you need it, give the material to work from, describe the shape of the answer (length, format, tone) and say what to avoid. Giving a reason helps more than a rule: "this will be read aloud, so no bullet points" lets the model handle cases you didn't think of.
Then treat the first answer as a draft. Point at what's wrong and ask again. Most good results come from 2 or 3 rounds, not one perfect prompt.
For example
"Summarise this" becomes "Summarise this contract for a tenant in 5 bullet points, flagging anything about deposits or ending early."
Put it to use
Where you'll meet prompt engineering in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.