A Clinical Reasoning + AI Calibration Coach

Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: A Clinical Reasoning + AI Calibration Coach

CAPABILITY GAP:
Clinical trainees increasingly encounter AI-supported decisions, but learning to use AI safely requires more than knowing what a tool can output. Learners must integrate foundational knowledge, uncertainty, patient context, and the possibility that the AI is wrong.

FIRST PROTOTYPE:
Create a clinical reasoning simulator in which an AI assistant sometimes offers useful, incomplete, or misleading suggestions. The learner must state an independent assessment, identify what evidence would change the decision, and explicitly answer 'what might I be missing?' before seeing expert feedback.

REPRESENTATIVE USE CASE:
A simulated patient case includes an AI-generated differential diagnosis that omits a less common but consequential possibility. The learner decides whether to accept, challenge, or investigate the suggestion.

LEARNING / HUMAN-SYSTEM FRAME:
Simulation with deliberate practice can support complex professional judgment when feedback is explicit and repeated. Metacognitive prompts make uncertainty and self-monitoring visible; the AI is deliberately treated as fallible evidence, not an authority.

QUESTIONS ALREADY IDENTIFIED:
1. How often should the AI be wrong to teach calibration without creating distrust?
2. Which measures capture reasoning quality rather than answer matching?
3. Does practice change behavior in authentic clinical settings?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to integrate AI advice into independent, evidence-based professional judgment, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: trainee, supervisor, patient case, clinical knowledge, AI assistant, institutional policy, and safety constraints. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: diagnostic reasoning, appropriate challenge of AI, uncertainty calibration, transfer, safety-critical misses, and explanation quality. A key disconfirming signal is: learners learn to distrust or obey the AI categorically instead of calibrating trust to evidence.