A recent study, published in the journal JAMA Network Open, revealed that the AI chatbot ChatGPT outperformed a group of doctors in diagnosing medical conditions based on case histories. The doctors, who had access to both conventional medical resources and ChatGPT, performed only slightly better than those who did not use the chatbot. The study highlighted two key findings: first, doctors often remained convinced of their initial diagnosis even when ChatGPT suggested a different and potentially more accurate diagnosis. Second, the doctors struggled to fully utilise ChatGPT's capabilities, failing to take advantage of its ability to process complex medical information and provide comprehensive diagnostic analyses. This suggests that while AI chatbots hold significant potential as tools for diagnosis, further research is necessary to understand how to effectively integrate them into medical practice and overcome the challenges of doctor-AI collaboration.
Briefing Doc: A.I. Chatbots vs. Doctors in Diagnosing Illness
AI Chatbots in Medicine: An FAQ
MY Times
Original Study