@yury.ai: Send this to a doctor friend and ask what they think 👇 On October 22 the journal Science published a Microsoft study: its AI was tested on 304 of the hardest medical cases, the kind doctors study as puzzles. It works like a panel of 5 virtual doctors, one proposes diagnoses, one orders tests, one hunts for mistakes, one keeps costs down and one double-checks it all. The result: the AI got the right diagnosis 83% of the time, general doctors with reference books 36%, and the best doctor in the study 52%. On everyday ER cases it’s ahead too, 80% versus 47% for doctors. The caveats matter: the AI competed against general doctors, not specialists, and it was tested on written case descriptions, not real patients. Follow @yury.ai #ai #artificialintelligence #medtech #microsoft #healthtech

172

2026-10-28 04:41:41

To see more videos from user @yury.ai, please go to the Tikwm homepage.