What is Prompting?

Prompting is how we tell a helpful computer what we want it to do. Imagine talking to a friendly robot: the words you give the robot are the “prompt.” A good prompt helps the robot understand you and give a useful answer.

Why prompting matters (simple):

• It is the main way to speak to language models.
• Good prompts make answers clearer, faster, and cheaper to get.
• In apps, prompts become part of how the product behaves — so we want them to be reliable.

Think of a prompt like a tiny recipe:

• Who the cook is (role)
• What to make (task)
• How it should look or taste (constraints and format)
• A picture of a finished dish (example)

Easy steps to write a prompt:

1. Say who the model should be. Example: “You are a friendly teacher.”
2. Say the task. Example: “Explain what a prompt is in two short sentences.”
3. Say how you want the answer. Example: “Use words a 10‑year‑old would know.”

Quick example you can try right now:

“You are a friendly teacher. Explain what a prompt is in two short sentences for a 10-year-old.”

If the model’s answer is too hard, add more rules like “use simple words” or “give one short example.” Small changes like this make big improvements.

Cheat sheet: Role → Task → Constraints → Example → Context. Use that order to keep prompts simple and reliable.