Prompt, context, loop

Three ways of working with AI. They build on each other, so it helps to be clear on what each one means.

Just starting

The ask

Prompt

What you ask for.

'Plan my week.' It works, but the answer is generic - it knows nothing about your week.

Building confidence

What you give it

Context

What you hand it to work with.

Same ask, but now it knows Thursdays are chaos and whose turn it is. The answer's yours.

Ready to build

The system that remembers

Loop

What makes it accumulate.

Tell it what was off, once. Next week starts sharper. It stops resetting to zero.

Each one contains the last. The loop doesn't replace the prompt - it runs it again, with memory.

A prompt is what you ask for. "Plan my week" is a prompt. It's the simplest way to use AI, and it's where most people start. The catch is that the tool only has your words to go on, so the answer comes back generic.

Context is the information you give it along with the ask. If the tool knows your Thursdays are busy, who does the school pickup, and which evenings you keep free, the same prompt gives you a plan that actually fits. The ask is the same; the result is better because you gave it more to work with.

A loop is when you keep using the same tool and it gets better over time. Instead of starting from scratch each time, you tell it what wasn't quite right, and it remembers for next time. The more you use it, the more it learns how you like things done.

Each one builds on the one before. Context makes a prompt more useful. A loop is a prompt and its context used again and again, with your feedback added each time. You don't stop using prompts; you just get more out of them.

Two things decide how much you get out of this: how clearly you can say what you want, and how well you can tell when it's not right yet.

The loop is the part worth aiming for, and it's what the build map walks you through.

The build map

The prompts