Your first reviewed AI project

What you’ll be able to do

• Define a bounded AI-assisted project with a useful deliverable.
• Test the result against explicit quality criteria.
• Document evidence, limitations, and the decisions you made.

Choose a small project with a real purpose

Your project should help someone do something concrete. Choose one of these options or design an equivalent low-risk task:

• Create an introductory workshop invitation and organizer checklist from fictional planning notes.
• Turn fictional meeting notes into a reviewed action list and brief.
• Compare two fictional workshop sessions using clearly defined metrics.

Do not begin with a task involving sensitive personal records, consequential decisions about people, or automatic external actions. A small project can demonstrate the full method without those complications.

Use a complete project brief

Write the audience, input sources, intended output, permitted tool use, and review process. State what success means before generating anything. Include one thing the system must never invent and one condition that should cause it to stop.

For example: “The invitation must use only confirmed dates and locations. It must label a missing time as pending. A human must approve the final text before publication.”

A worked example: workshop planning packet

The packet has three parts: a 100-word invitation, a confirmed-facts list, and an organizer checklist. Its fictional source mentions the library, Saturday as a proposed date, and an unconfirmed start time.

A strong packet keeps Saturday provisional, flags the missing time, and makes no claim about instructor credentials. The checklist assigns no owner unless the notes name one. The reviewer can trace every factual statement back to its source.

A packet with polished wording but an invented registration link does not pass the standard, even if most of the rest is useful.

Try a planning prompt

Help me plan a small AI-assisted project, using fictional or public information only.
Ask me for the audience, source material, deliverable, and success criteria before drafting.
Separate confirmed facts from assumptions.
Propose three test cases: a normal input, a missing-information input, and a conflicting-information input.
Identify the human review step and a fallback if a test fails.
Do not send, publish, book, or change anything in an external system.

Your practice: build, test, and revise

1. Write your brief and three quality criteria. Include factual accuracy and audience fit.
2. Prepare a source packet that you are allowed to share. Label fictional material clearly.
3. Create a first draft with AI assistance and preserve the prompt.
4. Test normal, missing, and conflicting inputs. Record the outputs and failures.
5. Correct the draft and revise the instructions where useful. Rerun all three tests.
6. Review privacy, unsupported claims, and barriers to participation.
7. Produce a final artifact and a short project note explaining what changed.

Your finished artifact: The final deliverable, source packet, prompt, test log, and a 200-word project note. Another person should be able to understand how you reached the result without trusting the assistant’s confidence.

Use this review rubric

Score each item 0 for missing, 1 for partly met, or 2 for met:

• Traceability: Important facts connect to identifiable source material.
• Task fit: The output serves the audience and stated purpose.
• Boundaries: Missing information remains explicit and permissions are respected.
• Review: Calculations, factual details, and substantive revisions have been checked.
• Repeatability: Your brief and tests allow someone to repeat the process.

This ten-point rubric supports reflection; it is not a certification. A high total does not excuse a serious factual or privacy failure. Correct such a failure before using the artifact.

Share thoughtfully

If you share the project, explain where AI assisted and where you made the decisions. Follow the disclosure rules that apply to the setting. Ask for feedback on the result and the method, then identify one change for a second version.

Reflection

What can you now do that you could not do before the course? Point to evidence in your project rather than describing confidence alone.