As American society continues its rapid adoption of artificial intelligence (AI), physician-scientists and students at Yale School of Medicine (YSM) are leveraging the use of this technology across multiple disciplines. For insights into AI’s role in advancing discovery and innovation at YSM, we spoke with Nancy J. Brown, MD, the Jean and David W. Wallace Dean of Yale School of Medicine and C.N.H. Long Professor of Internal Medicine.
How is Yale School of Medicine uniquely positioned to maximize the benefits of AI?
Yale School of Medicine has a long history in informatics that was distributed throughout the school, and we've now brought that together in a single academic home—the Department of Biomedical Informatics and Data Science. Because our informatics was distributed, there's a wonderful hub-and-spoke model and grass roots effort that's been taking place in AI, involving both the basic sciences and the clinical departments.
What are some of the most impactful ways that the school is currently using AI?
We are using AI to do computational biology to develop new ways to analyze very complex data coming from single-cell expression, for example, which is allowing us to make fundamental discoveries about how molecules work, how cells work, and potentially then to identify druggable targets. At the other end of the spectrum, we're using AI in the clinical arena to detect disease, often before we can identify it clinically. It’s very exciting stuff!
Is there a particular moment that allowed you as a physician-scientist/dean to fully grasp the potential of AI?
In the fall of 2023, the Provost’s Office organized a field trip to the Bay Area for members of the university cabinet. We spent several days talking with experts engaged in this work—focusing on what is the role for academia in AI versus industry, and how could we complement each other? When we got back, the Provost’s Office organized a series of faculty panels, which then presented to the Yale Task Force on AI. All the deans attended other schools’ faculty panels, which identified cross-cutting themes. That whole process really energized the university about AI.
What foundational principles are being applied in the school’s use of AI?
FAIR [Findability, Accessibility, Interoperability, and Reusability] principles. For example, when large language models use patient data, often those models are trained on very homogeneous populations, resulting in findings that may not be applicable to other groups. Because of our diverse patient population in New Haven and across Connecticut, our faculty has the ability to test and validate models and develop new models in populations that are reflective of the population of the United States.
How is multidisciplinary collaboration advancing Yale School of Medicine’s adoption of AI?
When you think about AI in medicine or in science, you need teams that are trained in biology and have domain expertise as well as those that are trained in AI or know how to think about AI. One of the outcomes of the task force meetings was the appreciation that our graduate students in the biological sciences need to be educated in AI, and our graduate students in computational science need to be educated in biology.
What excites you most about the school’s use of AI?
What excites me the most, [speaking] personally, is that this work is being driven by our young, smart faculty.