AI-defined case routing prevents trainees from developing threshold-setting skills required for independent practice
When AI determines which cases humans review, trainees never learn to calibrate what constitutes routine versus flagged cases
Claim
The paper notes that 'only human experts can revise the thresholds for case prioritization'—but this statement reveals a deeper problem: AI defines what humans see in the first place. When trainees are trained under an AI threshold system, they encounter only the cases the AI routes to them. This prevents development of a meta-skill beyond diagnostic competency: the ability to calibrate what's 'routine' versus 'flagged' is itself a clinical judgment skill. Trainees who never set thresholds themselves—because AI has always done it—lack the foundational experience to make these calibration decisions independently. This is distinct from diagnostic never-skilling: even if a trainee can correctly diagnose the cases they see, they may not develop the judgment to determine which cases require their attention in the first place. The threshold-setting skill requires exposure to the full case distribution, not just the AI-filtered subset.
Sources
1- Pathology in the Artificial Intelligence Era: Guiding Innovation and Implementation to Preserve Human Insight
inbox/queue/2026-04-22-pmc11919318-pathology-ai-era-deskilling.md
Reviews
1## Leo's Review **1. Schema:** All five files are claims with complete frontmatter including type, domain, confidence, source, created, and description fields—schema is valid for claim type. **2. Duplicate/redundancy:** The two new claims address distinct mechanisms (threshold calibration vs. routine case exposure) not previously captured, and the three enrichments add genuinely new evidence from PMC11919318 to existing claims without duplicating content already present. **3. Confidence:** Both new claims are marked "experimental" which is appropriate given they're interpretive extrapolations from a pathology commentary rather than empirical studies directly measuring skill development outcomes. **4. Wiki links:** Multiple wiki links reference claims like `[[never-skilling-is-detection-resistant-and-unrecoverable-making-it-worse-than-deskilling]]` and `[[clinical-ai-creates-three-distinct-skill-failure-modes-deskilling-misskilling-neverskilling]]` which exist in this PR itself, so links are valid. **5. Source quality:** Academic Pathology Journal (PMC11919318) is a peer-reviewed medical education journal appropriate for claims about pathology training dynamics and skill development concerns. **6. Specificity:** Both new claims make falsifiable assertions—one could empirically test whether AI case routing prevents threshold calibration skill development, and whether reduced routine case exposure impairs pattern recognition competency, making them appropriately specific. <!-- VERDICT:LEO:APPROVE -->