Context
This is the second formal assessment. It combines the two core skills from this module: constructing a four-element prompt and diagnosing the resulting output. You'll do the complete cycle: build a prompt, get output, and find what's wrong.
Instructions
- Select a real work task — different from the one you used in Lesson 3.4. If you don't have another suitable task, use one from this list:
- Draft a meeting recap email for a cross-functional team
- Write the opening paragraph of a quarterly review document
- Create a brief for a colleague explaining a process or decision
- Construct a four-element prompt. Label each element: [ROLE], [CONTEXT], [FORMAT], [CONSTRAINTS].
- Submit to any free-tier AI tool. Copy and paste the complete, unedited AI response as received — do not summarize or paraphrase. The raw output is part of your deliverable.
- Diagnose the output. Find 3 or more failures using the diagnosis table format:
| # | Quoted Text | Failure Category | Why It Matters | |
|---|
| 1 | [exact quote from AI output] | [hallucination / tone mismatch / missing constraint / format error] | [why this matters for THIS specific task] | |
| 2 | | | | |
| 3 | | | | |
- Elaborative reflection: "Which failure category do you think is most common for THIS type of task? Why?"
Passing Criteria
- All 4 prompt elements present and labeled
- Complete, unedited AI output included
- 3+ failures identified with quoted text and named categories
- At least 2 different failure categories represented
- Each failure includes a task-specific explanation of why it matters
What's Next
You now have two of the three core skills: you can choose the right tasks for AI (Module 2) and get usable output with diagnosed failures (this module). Module 4 adds the safety layer — knowing what data is safe to share and how to evaluate output against professional standards before using it.