A.I. Chatbots Defeated Doctors at Diagnosing Illness (The New York Times)

A recent study led by Dr. Adam Rodman at Beth Israel Deaconess Medical Center revealed surprising insights into the use of ChatGPT-4 for medical diagnoses. While the AI chatbot outperformed physicians with an average score of 90% in diagnosing conditions from case reports, doctors using ChatGPT only slightly improved their scores (76%) compared to those without it (74%).

The study highlighted two key issues: physicians often dismissed AI suggestions that contradicted their own diagnoses due to overconfidence, and many failed to fully utilize the chatbot’s capabilities. Instead of leveraging its ability to analyze entire case histories, they treated it as a simple search engine. These findings underscore the potential for AI as a diagnostic tool but reveal a gap in training and trust among physicians to maximize its benefits.