Artificial intelligence (AI) has become embedded in medical practice and education. In today’s digital world, medical learners use AI tools to arrive at plausible diagnoses with speed and accuracy that can equal those of experienced clinicians. This shift challenges a long-standing assumption in medical education that clinical error primarily reflects gaps in factual knowledge. Digital information and AI now make facts immediately accessible. Errors arise when AI is misapplied, when users accept outputs with unwarranted confidence, and when clinicians fail at the therapeutic judgment required to act. Two brief outpatient encounters involving a third-year medical student illustrate the gap between technology-assisted diagnosis and the human decision to act. In both cases, the student used AI and digital resources to reframe the clinical problem in a useful way. The responsibility to verify the diagnosis, assess risk, and accept the consequences of action remained with the attending physician. AI collapses the distance between presentation and diagnosis. It leaves untouched the distance between knowledge and responsibility an intersection that defines medical professionalism and now focuses explicit attention in medical education.
Don K. Nakayama (Wed,) studied this question.