Prompting: Getting What You Want From AI
The one skill that improves every other thing you do with AI — what a prompt really is, how to build one, and what to do when the answer comes back wrong
Read it in the library →What a Prompt Actually Is
A prompt is not a search query. It is everything you hand over in one request — and understanding what the model can and cannot see explains almost every disappointing answer you have ever had.
The Anatomy of a Good Prompt
The parts that reliably help — perspective, context, constraints, outcome, format and worked examples — plus when showing beats telling and why negative instructions are weaker than positive ones.
Getting Output You Can Use
The gap between an impressive answer and a usable one — asking for structure, controlling length and tone, getting editable drafts, and making the model interrogate you before it answers.
When It Goes Wrong
Diagnosing a bad answer instead of retrying it — ambiguity, missing context, oversized tasks and plain error — plus the sycophancy trap and how to spot confident fabrication.
Prompting for Real Work
Turning one-off requests into reusable assets, sharing them across a team, supplying your own material safely, and recognising the point where a better prompt is no longer the answer.
Every lesson is free, with no sign-up. Reading happens in the library, where your progress is saved on your device.