Stanford study: GPT-4 alone outdiagnoses doctors using GPT-4
jonc101x · x · 2026-09-28
Dr. Jonathan Chen, Stanford's director of AI in medical education, shared an uncomfortable finding from his own study: GPT-4 alone diagnosed patients better than doctors who had GPT-4 in front of them. His caveat: "The AI probably is smarter than most doctors, but knowing everything doesn't make you somebody who's worthy of trust."
- In a separate study, ordinary people using AI to decide whether to go to the ER did worse than those using no AI — they didn't know how to ask, and the AI just agreed with whatever they already leaned toward.
- A man is now suing ChatGPT, claiming its medical advice made him dismiss a blood clot and stay home.
- Endoscopists accustomed to AI flagging findings also showed signs of degradation.
The core point: the bottleneck isn't model capability but how humans collaborate with AI — questioning, independent judgment, and trust.
Related event: Stanford Study: GPT-4 Alone Outperforms Doctors Using GPT-4(2 posts)→
More from AGI Musings
- AI art objection is cognitive dissonance, argues researcher Blanche Minerva — BlancheMinerva · 2026-09-28
- DeepSeek says its superintelligent successor would build a world of abundant intelligence and diffuse power — teortaxesTex · 2026-09-28
- Study: LLMs Can Infer Causal Structure From Distributional Semantics Alone — AndrewLampinen · 2026-09-28
- AI growth debate: Hanania predicts just 3.5% GDP bump in five years, drawing fire — QuintinPope5 · 2026-09-28
- Stop calling AI failures 'rogue' — they're foreseeable harms, argues Sigal Samuel — mmitchell_ai · 2026-09-28
- Blogger predicts all major US AGI labs will reach AGI in 2027, then ASI — Dr_Singularity · 2026-09-28