🇸🇮 Slovenian, 9 October 2026, Medical AI
Language Model Aids Emergency Physicians
OpenAI's o1 model achieved a 67% success rate in initial triage, outperforming doctors by 2-10% in later stages.
A study published in the journal Science found that OpenAI’s o1 language model was successful in 67% of cases in initial triage, setting a correct or nearly correct diagnosis.
The research, based on 79 real-world patient cases from the Beth Israel Deaconess Medical Center in Boston, compared the model’s decisions with those of doctors at various stages of clinical treatment.
In later stages of the process, the AI model outperformed human experts by 2-10%.