AI and medicine: Transforming competence, learning and practice

We stand at the beginning of an era quickly transforming with the evolution of machine learning and generative artificial intelligence (AI). In health care, computers have been indispensable since I was a medical student in the 1980s. I recall using the first computers for word processing, organizing my class notes and later my research data. With the evolution of presentation software in the 1990s, PowerPoint improved my academic presentations and internet communication allowed me to connect with colleagues around the world. When the pandemic struck, many physicians pivoted rapidly to virtual care and today I could not run my clinical psychiatry practice without MS Teams and Zoom for videoconferencing with students, colleagues and patients.

The technology behind these applications is evolving rapidly. The electronic patient record that I have used for two decades, today incorporates AI that can analyze my orders and warn me if I prescribe a problematic combination of medications for a patient with a particular diagnosis. New AI-enabled systems can listen in on patient encounters and combine this with patient chart, laboratory and imaging data to generate a clinical note and plan. Doctors and patients will soon have access to information about diagnosis and prognosis that is based on an entire health record, rather than just the data we have time to gather before our appointments and consultations.

Metacognition: a future core competence

In education, the collection of information about practice and learning will aid competence-based models, and learning analytics will provide new tools to identify learning disabilities and create learning growth curves. It didn’t take long for innovators to use generative AI to challenge a certification exam. The first version of ChatGPT passed the United States Medical Licensing Examination with a score of 60 per cent in early 2023 and just a few months later, Google’s Med-PaLM received 85 per cent and then Microsoft’s GPT-4, 86.7 per cent.[i] This development highlights our need to continue to move beyond examinations comprised of factual information toward high order interpretation of complex data.

Generative AI also raises new learning opportunities, particularly in its ability to create and analyze sophisticated text. Educators have begun to help students understand how generative AI accesses information and creates compelling arguments, in a way that recalls early versions of evidence-based medicine. A priority for the Royal College will be to identify the ways in which competence will be transformed in the future. A core competence for future medical practice, for example, will be metacognition — the ability to analyze and critique patterns and biases in both human and AI-based problem solving and how they interact with each other. Another critical ability will be understanding the biases that can arise when computer algorithms are used to generate diagnostic hypotheses and probabilities.

The risk of artificial intelligence is hotly debated and there have been calls for moratoria.[ii] But, it is likely already too late to stop this evolution; most people carry with them a smart phone that is already replete with apps utilizing artificial intelligence. All search engines use AI, as do shopping and banking apps and those used to book flights and hotels. When we “talk” to Siri or Alexa or Google we are engaging in a relationship with an AI algorithm. Further, almost everyone I know has experimented with ChatGPT or a similar generative AI system, to do everything from writing a limerick to creating a scientific keynote address.

Join the discussion: Forum on Generative AI in October

The Royal College is responding to this technological evolution and its myriad of applications to medical education and practice. Having commissioned both a Task Force on Research in 2018 and another focusing on Artificial Intelligence in 2019, the Royal College has been focused on implementing the recommendations.

Among these are the establishment of a Research and Evaluation Advisory Committee (REAC) that will be chaired by Richard Reznick, MD, FRCSC and Past-President of the Royal College, and coordinated by our associate director, Research and Evaluation, Tanya Horsley, PhD, MBA.

The Research and Evaluation Department, in concert with the Data and Informatics Governance Committee of Council, is hosting a panel on AI to discuss

  • emerging competencies,
  • uses of technology to propel learning and assessment, and
  • how the Royal College can help the medical profession prepare for the profound changes that AI will bring to education and practice.

The Forum on Generative AI will take place virtually on October 25, 2023, at 12:00 p.m. EDT. Dr. Horsley and I will be joined on the panel by Lorelei Anne Lingard, PhD, FRCPSC (Hon), professor and senior scientist at Schulich School of Medicine & Dentistry, Western University, and Michael Caesar, MBA, PfMP, chief data & analytics officer at the University Health Network. Bertalan Meskó, MD, PhD, The Medical Futurist and director of The Medical Futurist Institute in Hungary, will give an address. Learn more and register for the event.

No matter where we practice — be it in a rural community or a large urban centre, solo or group practice, and in any of the 72 Royal College specialties and subspecialities — our teaching, research and clinical activity is being transformed by technology. Our Royal College will be there as a partner to help us adapt to this new reality.


Brian Hodges, MD, FRCPC, PhD

Brian Hodges is the 47th President of the Royal College of Physicians and Surgeons of Canada. He is the executive vice-president of education and chief medical officer at the University Health Network. A professor in the University of Toronto’s Temerty Faculty of Medicine and at the Dalla Lana School of Public Health, as well as a senior fellow at Massey College, Dr. Hodges is also a practising psychiatrist.

[i] Loh E. BMJ Leader 2023;0:1–4. doi:10.1136/leader-2023-000797

Mbakwe AB, Lourentzou I, Celi LA, et al. ChatGPT passing USMLE shines a spotlight on the flaws of medical education. PLOS Digit Health 2023;2:e0000205.

[ii] New York Times (2023), and Future of Life Institute (2023),