AI doctors still lag real physicians on safety
A new benchmark using real patient-doctor chats reveals AI models still make more unsafe errors than human physicians.
A new benchmark using real patient-doctor chats reveals AI models still make more unsafe errors than human physicians.