14 min · up to 75 XP

Be clear and direct

The golden rule of prompting, with before/after rewrites.

Priya's team have caught the AI habit, and the team channel has caught a matching complaint: "it just gives me waffle." She scrolls back through what everyone has actually been sending. Exhibit A: Aisha, from the customer team, asking about the spring newsletter.

Read it the way you learned in lesson one, as a prediction engine would. It doesn't judge your ask; it completes it. So before Priya says anything: what comes back?

Predict the response

Aisha wants help with the spring newsletter draft. Here's her message to the AI, verbatim.

Hi! Sorry to bother you, I was just wondering if you could maybe have a quick look at our newsletter and share some general thoughts at some point? No worries if it's too much!

What do you think the AI will do?

Look at what Aisha's message actually tells the model. Maybe. Quick look. Some general thoughts. At some point. Every hedge widens the range of acceptable replies, and a prediction engine fills a wide target with the most average text that fits.

Being direct isn't being rude. "Please" and "thanks" cost nothing. The damage comes from vagueness dressed as politeness: softening the deliverable until there isn't one. Try the two fixes yourself and watch the reply change.

This lesson is part of the full course

The free lessons are just the opening chapter. Unlock the rest of AI at Work to keep going.

Every AI at Work lesson, as it ships
New interactive lessons added monthly
Founding-member pricing, locked in
Learn at your own pace, forever

or keep browsing the free lessons