Once again, a study shows that AI often makes mistakes with atypical patients or incomplete information. Patients, as was known, then trust the AI too much. Unfortunately, the study also shows the same behavior among doctors. More on counter-strategies later; the problem is not unsolvable. (translated)


