Discernment: The AI Fluency Competency Behind This Entire Task Statement
CoreEvaluate output accuracy and completeness · Difficulty 2/5
Explanation
Naming the Skill This Task Statement Has Been Teaching
Everything in this task statement -- setting success criteria before reading a draft, running the five-part checklist, catching completeness dressed up as thoroughness, catching relevance dressed up as a nearby answer -- is a single named competency in Anthropic's AI Fluency Framework: Discernment, the discipline of judging output against what was actually asked for, what the sources actually say, and what your field considers acceptable. Success criteria and the checklist are not separate tricks; they are Discernment made concrete and checkable.
Discernment vs. Diligence: How vs. Why
This course names *Diligence* explicitly elsewhere (iteration in lesson 1.3, ethical use in lesson 6.4) as the competency of staying responsible for an outcome and using AI effectively, ethically, and safely. Discernment is Diligence's quieter partner, and the exam expects you to tell them apart cleanly:
| Discernment | Diligence | |
|---|---|---|
| Answers the question | *How* do I review this well? | *Why*, and *when*, must I review at all -- and who owns the result? |
| Shows up as | The five-part checklist, checkable success criteria, tracing a claim to its source | The decision to actually run the review, escalate to a human, or disclose AI use, even under time pressure |
| A failure looks like | Missing a dropped requirement because the checklist wasn't actually applied | Skipping the checklist altogether because the deadline is close and the draft "looks fine" |
| Where else it's named | Only here (2.1), though it's exercised throughout Domain 2 | Explicitly in Lessons 1.3 and 6.4 |
A useful compression: Discernment supplies the method for reviewing well; Diligence supplies the reason the review has to happen at all, and names who answers for it if it didn't.
How the Two Compound in Practice
The two competencies don't operate independently -- they compound, and the exam's trickiest scenarios exploit the gap between having one without the other. Discernment without Diligence looks like an Associate who knows the five-part checklist cold but, rushed before a deadline, skips running it and ships the draft anyway -- the skill existed, the discipline to apply it under pressure didn't. Diligence without Discernment looks like an Associate who dutifully reviews every output every time, but only checks tone and length, missing that a citation was fabricated, because they never learned to apply the actual checklist -- the discipline existed, the skill to execute it well didn't. Both failures produce the same visible result (an error reaches production), but the fix is different in each case: the first needs a workflow gate that forces the review to happen regardless of time pressure; the second needs training on what to actually look for. Recognizing which one is missing is itself part of Discernment applied one level up -- discerning the gap in your own process.
A Self-Check Heuristic for Exam Scenarios
When a scenario describes an evaluation failure, ask: did the reviewer *not know what to check* (a Discernment gap -- they lacked the checklist, the success criteria, or the source-tracing skill), or did they *know what to check but not check it* (a Diligence gap -- time pressure, complacency, or treating the task as beneath verification)? A scenario that hinges on "they didn't realize completeness and relevance are separate failure modes" is testing Discernment. A scenario that hinges on "they knew better but skipped the check because the deadline was tight" is testing Diligence.
Common exam traps
- Treating Discernment and Diligence as synonyms because both concern "careful use of Claude." They're distinct competencies answering different questions -- how to review well, versus why and when review is non-negotiable.
- Assuming a scenario about missing a checklist step is automatically a Diligence question. If the reviewer genuinely didn't know to check for it, that's a Discernment gap, not a discipline lapse.
- Forgetting that having one competency doesn't guarantee the other -- a skilled reviewer can still skip the review (Diligence gap), and a disciplined reviewer can still review poorly (Discernment gap).
Key Takeaways
- Discernment is the AI Fluency competency for judging output against what was requested, what the sources say, and field standards -- it's the named skill behind the success criteria and five-part checklist taught in this task statement
- Discernment answers 'how do I review well'; Diligence (named explicitly in Lessons 1.3 and 6.4) answers 'why and when must I review, and who owns the result'
- You need Discernment (skill) to review well and Diligence (discipline) to actually review every time, even rushed -- one without the other still produces failures
- Self-check for exam scenarios: a reviewer who didn't know what to check has a Discernment gap; a reviewer who knew but skipped the check has a Diligence gap
Glossary Terms
Related Concepts
Defining Success Criteria Before Evaluating Output
Decide what 'good' looks like before evaluating an output, not after
The Accuracy, Completeness, Relevance, Consistency, and Audience-Fit Checklist
The five-part checklist is accuracy, completeness, relevance, consistency, and fitness for audience
Capability Hallucination: Claiming an Action It Never Took
Capability hallucination is Claude claiming to have taken an external action (emailed a file, saved a document) it never actually took