Module 01 Activity

Activity ID: PAW-M01-A01 Estimated active time: 30-40 minutes

Purpose

This activity turns unclear requests into task briefs you could hand to a colleague. The test is not whether the AI output improves. It is whether another person could tell what good would look like.

Steps

1. Take three requests you have actually made, or three safe fictional ones.
2. For each, write the job being done in one sentence, without naming a tool.
3. Name the reader and what they do after reading.
4. State what the output must contain, avoid, and preserve.
5. Write one sentence on what would make the answer unusable.
6. Give the three briefs to someone else and ask what they think the output will look like.

Evidence to hand in

• Three original requests, unedited.
• Three task briefs.
• What your reader predicted for each.
• One line per brief on what you would change.

Review checklist

• Each brief names the job without naming a tool.
• Each names a reader and a decision.
• Each says what would make the answer unusable.
• Someone else predicted the output correctly from the brief alone.

Mark complete