---
id: "claim-hinton-radiology-error"
type: "claim"
source_timestamps: ["¶1", "¶2", "¶3"]
tags: ["historical-analogy", "ai-predictions"]
related: ["concept-induced-demand", "concept-complementarity", "entity-geoffrey-hinton"]
confidence: "high"
testable: true
speakers: ["Chengwei Liu", "Balázs Kovács"]
sources: ["futures"]
sourceVaultSlug: "hbr-seg-futures"
originDay: 2
articleStem: "hbr-cl-84-big-tech-capability-crisis"
sourceUrl: "https://hbr.org/2026/06/big-techs-looming-capability-crisis"
sourceTitle: "Big Tech’s Looming Capability Crisis"
---
# Hinton's Radiology Prediction Failed Due to Basic Economics

## Claim: Hinton's Radiology Prediction Failed Due to Basic Economics

**Confidence: high · Testable: yes**

[[entity-geoffrey-hinton|Geoffrey Hinton's]] 2016 prediction — that deep learning would replace radiologists within **5–10 years** — was wrong. The authors argue it failed **not because the technology failed**, but because Hinton ignored [[concept-induced-demand|induced demand]] and [[concept-complementarity|complementarity]].

By 2025, radiologist pay hit **$570,000** with severe shortages (**130 days to fill** a role). AI made imaging cheaper, which *expanded* the market and made the human sign-off *more* valuable, not less.

> Enrichment: The prediction is accurately described, but the "entirely wrong" framing is **interpretive rather than strictly empirical**. A counter-reading: AI has *shifted* radiology work, triage, and image interpretation rather than eliminating radiologists — so "failed prediction" may be too absolute, and the case may not generalize cleanly to software engineering (which is globally scalable and less regulated).


## Related across articles
- [[claim-professional-services-disruption]]
- [[concept-complementarity]]