OpenAI's o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors (theguardian.com)

• AI diagnosed 67% correctly vs. doctors' 50-55%
• Study conducted in an ER using clinical data
• Model designed to assist, not replace, physicians

"OpenAI's o1 model outperformed triage doctors in an emergency room study, correctly diagnosing 67% of patients compared to the physicians' 50-55% accuracy. The AI used clinical notes and lab results, showing significant potential for reducing misdiagnosis. Researchers stressed it is a support tool, not a replacement for human doctors."