#ACGME2024 Session Summary: The Generative AI Revolution and Innovations and Opportunities in Medical Education

March 25, 2024

A room full of seasoned medical education professionals gasped, then stunned into silence, as a disembodied voice from a smartphone calmly described how to deal with a resident in an empathetic way to address a list of shortcomings – just seconds after being presented with a difficult, multi-part scenario. For many in the audience, this was a demonstrative example of the power of generative artificial intelligence (AI), leaving some bewildered and others wanting more.

For the next hour and a half, four medical education leaders took the audience, attendees of the 2024 ACGME Annual Educational Conference's Featured Plenary “The Generative AI Revolution: Innovations and Opportunities in Medical Education,” on an AI journey – how we got here, how quickly it is evolving, and how to harness it in a way that complements medicine and medical education, while also addressing its many challenges.

Putting his smartphone away and bringing the room of medical educators to a sobering reality, Robert M. Wachter, MD explained, “your [learners] are already doing this, and are getting pretty good at it.”

Dr. Wachter is professor and chair of the Department of Medine at the University of California, San Francisco, and a bestselling author and recognized expert on technological opportunities and challenges. He discussed the history of AI, and how earlier missteps – focusing on diagnosis (the “hardest” problem) – caused the medical world to reapproach the issue slower and more cautiously than in other sectors. However, it has quickly become a gamechanger, and he foresees the opposite progression now, by starting with lower stakes areas such as paperwork before evolving to higher stakes components like diagnostics.

Dr. Wachter urged human vigilance and preventative steps to protect against what he described as “deskilling.” AI in medicine in the foreseeable future, even with the breakneck pace of evolution, he said, will be useful but not fully trustworthy. He warned, “humans suck at vigilance… if they are not actively engaged.”

As someone already employing AI in his work, panelist Jesse Burk-Rafel, MD, MRes, the assistant director of the Precision & Translational MedEd Lab and assistant professor of internal medicine at New York University Grossman School of Medicine, called AI the “great leveler” for those with more limited resources. He addressed the use of AI in areas such as admissions, assessment, simulation, and even self-directed learning.

Medical education is behind in AI, argued panelist Eva Aagaard, MD, FACP, professor of medical education, vice chancellor for medical education, and senior associate dean at Washington University School of Medicine. She urged medical educators to think about how to incorporate AI into education and training, but with consideration to risk and benefit of use, how to gauge accuracy of AI output, and even basic AI literacy. Importantly, Dr. Aagaard noted the importance of ethics in this area, as well as the need to address how AI could affect the professional identity of physicians.

Eric Holmboe, MD, MACP, FRCP, currently the chief executive officer of Intealth, and until the end of 2023, the chief research, milestone development, and evaluation officer at the ACGME, focused on AI and assessment. He noted that tools are already being developed for AI to help identify struggling learners early in the process and cautioned against bias in AI and the need for vigilance. AI should be an enhancement, but humans must remain in the loop, he emphasized.

Finally, one of the questions closing the session pondered whether AI would render certain specialties obsolete in practice. The panelists steered the response in another direction, with Dr. Holmboe further arguing for ethics in AI’s relationship with medicine, especially on bias, and Dr. Aagaard stressing the need for human vigilance to prevent such problems.

“What is the critically, uniquely human element that AI cannot replace?” asked Dr. Burk-Rafel. Education, he said, can utilize AI in a way that is safe and fair, honoring the social contract.

Dr. Wachter suggested that where medicine and medical education are now with AI is reminiscent of something Ernest Hemingway wrote in The Sun Also Rises, that it’s coming “gradually, then suddenly.”

“It is here,” added Dr. Holmboe, “[and] it can really make a difference.”