The question may sound closer to science fiction, but today it is being seriously raised within one of the most prominent medical journals in the world: JAMA. The concept that has sparked global debate can be summarized as follows: If artificial intelligence has become capable of reading medical tests, analyzing X-rays, suggesting diagnoses, and participating in medical decision-making, is it enough to treat it merely as 'computer software'? A number of top experts in medicine and technology ethics believe the answer is no longer that simple. Instead, they propose a future where medical AI could undergo a path identical to that of a human physician. This would include exams similar to the United States Medical Licensing Examination (USMLE), specialized tests like the Board, and a supervised 'residency training period' under the oversight of doctors before being permitted broad clinical participation. Most interestingly, the proposal does not speak of a traditional technical approval—as is the case with current applications and programs—but rather of a time-limited 'practice license' that requires continuous renewal, performance monitoring, and evaluations for errors and medical safety. In other words, we may one day reach a stage where medical artificial intelligence is treated more like a 'supervised intelligent medical assistant' rather than just an electronic program. Yet, despite this stunning development, the most critical question remains: Can a machine understand a human the way a doctor does? Medicine is not just numbers and diagnoses; it is experience, intuition, responsibility, humanity, and the reassurance given to a patient who sometimes seeks a word of comfort before medicine. Perhaps for this reason, and despite all the progress, the world is still speaking of 'artificial intelligence under the supervision of a doctor' and not 'instead of a doctor.' JAMA Opinion Article / HealthExec
The question may sound closer to science fiction, but today it is being seriously raised within one of the most prominent medical journals in the world: JAMA. The concept that has sparked global debate can be summarized as follows: If artificial intelligence has become capable of reading medical tests, analyzing X-rays, suggesting diagnoses, and participating in medical decision-making, is it enough to treat it merely as 'computer software'? A number of top experts in medicine and technology ethics believe the answer is no longer that simple. Instead, they propose a future where medical AI could undergo a path identical to that of a human physician. This would include exams similar to the United States Medical Licensing Examination (USMLE), specialized tests like the Board, and a supervised 'residency training period' under the oversight of doctors before being permitted broad clinical participation. Most interestingly, the proposal does not speak of a traditional technical approval—as is the case with current applications and programs—but rather of a time-limited 'practice license' that requires continuous renewal, performance monitoring, and evaluations for errors and medical safety. In other words, we may one day reach a stage where medical artificial intelligence is treated more like a 'supervised intelligent medical assistant' rather than just an electronic program. Yet, despite this stunning development, the most critical question remains: Can a machine understand a human the way a doctor does? Medicine is not just numbers and diagnoses; it is experience, intuition, responsibility, humanity, and the reassurance given to a patient who sometimes seeks a word of comfort before medicine. Perhaps for this reason, and despite all the progress, the world is still speaking of 'artificial intelligence under the supervision of a doctor' and not 'instead of a doctor.' JAMA Opinion Article / HealthExec
The question may sound closer to science fiction, but today it is being seriously raised within one of the most prominent medical journals in the world: JAMA. The concept that has sparked global debate can be summarized as follows: If artificial intelligence has become capable of reading medical tests, analyzing X-rays, suggesting diagnoses, and participating in medical decision-making, is it enough to treat it merely as 'computer software'? A number of top experts in medicine and technology ethics believe the answer is no longer that simple. Instead, they propose a future where medical AI could undergo a path identical to that of a human physician. This would include exams similar to the United States Medical Licensing Examination (USMLE), specialized tests like the Board, and a supervised 'residency training period' under the oversight of doctors before being permitted broad clinical participation. Most interestingly, the proposal does not speak of a traditional technical approval—as is the case with current applications and programs—but rather of a time-limited 'practice license' that requires continuous renewal, performance monitoring, and evaluations for errors and medical safety. In other words, we may one day reach a stage where medical artificial intelligence is treated more like a 'supervised intelligent medical assistant' rather than just an electronic program. Yet, despite this stunning development, the most critical question remains: Can a machine understand a human the way a doctor does? Medicine is not just numbers and diagnoses; it is experience, intuition, responsibility, humanity, and the reassurance given to a patient who sometimes seeks a word of comfort before medicine. Perhaps for this reason, and despite all the progress, the world is still speaking of 'artificial intelligence under the supervision of a doctor' and not 'instead of a doctor.' JAMA Opinion Article / HealthExec
comments
Could AI one day hold a specialization certificate or a "license to practice" like a doctor?
display options :
Full
Main image only
without images
comments