WebMD, And The Tragedy Of Legible Expertise

A census standout. WebMD is useless (everything reads as cancer; aspirin and the far-deadlier warfarin get indistinguishable scary side-effect lists) precisely because it is too big, legible, and canonical — it must placate every stakeholder and dodge every lawsuit, so it optimizes for not-getting-sued rather than for being useful (Moloch). Scott's own small psychiatry database is better only via 'security by obscurity.' He then generalizes: Dr. Fauci is 'the WebMD of people' — legibly okay; Zvi Mowshowitz is illegibly great because he optimizes for one thing (being right), but you can't put Zvi on the throne (he'd either learn to play politics and lose his edge, or be forced out). The CDC Director must optimize for two things — being right AND keeping power — so on a harder problem she predictably underperforms Zvi despite more resources. The exported frame: legibly-mediocre experts vs. illegibly-great individuals; the expert system is 'a machine that takes Moloch as input and produces Anthony Fauci as output,' astonishing and underappreciated. 'Should we trust experts? Depends who you mean by we' — Scott can follow Zvi via illegible recognition, but that doesn't scale (others would trust their MAGA-cap friend). Rejects 'Marx's Fallacy' (burning down the system won't elevate the non-power-optimizers) in favor of gradual improvement — 'but prediction markets would still be better.'

Quality 81 · Excellent. High-Excellent; a genuine standout scored on merit. The 'legible vs illegible expertise' framing is a foundational, frequently-cited contribution, and the WebMD->Fauci->Zvi progression plus the 'Moloch-to-Fauci machine that's better than we deserve' synthesis resolve the trust-the-experts question with rare balance (neither expert-worship nor expert-burning). Co-tier with Missing-Heritability (81), just below the Bounded-Distrust / Cult-of-Smart cluster (82). Tail-down does not apply to a real standout essay. 81.

Claude’s paradigm shift 58 · Moderate. Top of Moderate / Notable. A novel application of James C. Scott's 'legibility' (and Scott's own Moloch) to expertise and institutions; 'legible vs illegible expertise' genuinely entered the discourse as a new lens, though it builds on existing concepts rather than inventing one whole. 58.

Real-world impact 3 · Moderate. The 'legible vs illegible expertise' framing (WebMD → Fauci → Zvi) became a foundational, frequently-cited contribution that resolves the trust-the-experts question with rare balance. Real conceptual influence within rationalist/epistemics discourse, but no material or population-level change — modest RWI.