Module 02: AI and Prompting Foundations
Lesson 02.5: How to Iterate Instead of Starting Over
Lesson Promise
Improve Copilot outputs through focused revision prompts.
Real-World Scenario
A learner discards every imperfect first answer instead of asking Copilot to fix specific weaknesses.
Core Concept
The first answer is usually a draft. Expert users improve it with targeted follow-ups: clarify, shorten, compare, verify, restructure, or change tone.
Iteration works best when you tell Copilot what succeeded, what failed, and what standard the next version must meet.
Step-By-Step Workflow
- Review the first answer.
- Identify one weakness.
- Ask for a targeted revision.
- Request assumptions or missing facts.
- Compare versions.
- Save the best reusable prompt.
Prompt Lab
Bad Prompt
Try again.
Better Prompt
Revise this to be shorter, more specific, and clearer about next steps. Keep the same facts.
Expert Prompt
Review your previous answer against these criteria: accuracy, source support, completeness, tone, actionability, and risk. Then revise it and explain what changed.
Hands-On Exercise
Improve one Copilot output through three targeted follow-ups.
Deliverable
An iteration log with before/after versions.
Prompt Foundations Checklist
Common Mistakes
- Asking vague prompts and blaming the tool for vague answers.
- Providing context without source boundaries.
- Forgetting to specify audience, format, tone, and length.
- Restarting every time instead of iterating.
- Skipping review because the answer is well written.
Quiz / Checkpoint
What should an iteration prompt name?
The specific weakness, the desired change, the facts to preserve, and the review criteria.
