← sources

An AI tool just performed four times better than physicians in diagnosing, a recent study shows

source: reddit (reddit:ChatGPT)published 2026-10-23 03:38ingested 2026-10-24 02:36

retained text

[Here's the peer reviewed study](https://www.science.org/doi/10.1126/science.aeh7930) that shows the AI tool being **FOUR TIMES** more accurate than human physicians - the AI tool achieved an accuracy of 80%, while the physicians averaged 20%.

*^(For those who are not familiar with)* [*^(Science.org)*](http://Science.org)*^(, Science is one of the world's top academic journals and employs a rigorous peer-review process for all of its research articles. It's one of the most prestigious and selective journals in the world at an acceptance rate of less than 8% of all submissions. To be published there, a paper must typically demonstrate a significant leap in knowledge or a major methodological breakthrough and pass a stringent peer-review process where independent experts in the field scrutinize the data, logic, and conclusions. So basically, the vetting for this study is among the most rigorous available in academia.)*

[\[The Reuters article\]](https://www.reuters.com/business/healthcare-pharmaceuticals/microsoft-develops-ai-tool-that-can-diagnose-sequentially-like-doctor-2026-10-23/)

document fetched 2026-10-24 00:21 · canonical https://reddit.com/r/ChatGPT/comments/1xd3cs3/an_ai_tool_just_performed_four_times_better_than