Robert M. Wachter, MD, opened Kaiser Permanente Bernard J. Tyson School of Medicine’s (KPSOM) 2026 Curriculum Retreat with a clear warning and an invitation. Artificial intelligence, he said, will be “the most important issue in medicine and in medical education for the foreseeable future,” and schools that fail to embrace and understand it “will not be great in the future.” His keynote address became the lens through which faculty, staff, students, and leaders spent the day examining how AI is transforming healthcare and what that means for KPSOM’s curriculum.
Dr. Wachter, chair of the Department of Medicine at UC San Francisco and author of A Giant Leap: How AI Is Transforming Healthcare and What That Means for Our Future, described his view on AI as “informed optimism.” AI’s promise in healthcare, he argued, stems from two forces: rapidly advancing capabilities and deep, longstanding gaps in access, quality, and clinician well‑being. Together, they make this a pivotal moment for medical schools.
“I think [KPSOM has] a tremendous advantage, being still relatively new, as you think about this,” Wachter told the audience. “And I'm glad you're taking the day to think hard about it. I personally do think this is going to be the most important issue in medicine and in medical education for the foreseeable future … it's going to be transformative.”
Wachter traced AI’s evolution since the public release of ChatGPT in 2022, when it was initially known for doing tasks such as writing poems and short essays. Large language models now pass licensing exams, perform impressively on complex diagnostic puzzles, and in some studies matching or exceeding physicians’ input on tough cases, he said. In controlled scenarios, patients sometimes rate AI‑generated responses as more empathetic and better communicated than those of clinicians.
At the same time, Wachter stressed AI’s limitations. Systems may still produce false answers and can amplify bias and misinformation. Studies have shown that clinicians, acting as the “doctor in the loop,” sometimes worsen AI’s performance rather than improve it. And as technologies take over more tasks, educators must confront the risk of “deskilling” clinicians in areas where human judgment remains essential.
Wachter argued that newer schools like KPSOM have “a massive strategic advantage” over legacy institutions. Older organizations are often “too fixed to [their] prior way of thinking about the world” to change quickly. KPSOM, by contrast, has the opportunity to integrate AI “into [its] DNA” from the start.
Dean and CEO John L. Dalrymple, MD, echoed that message in his welcome remarks. The school, he reminded participants, was created to train “physicians of the future” who can expect, adapt to, and lead change. Developing “a comprehensive and coherent approach to AI at our school … is no longer optional. It is essential,” Dalrymple said. Borrowing a hockey metaphor, he urged the community to “skate to where the puck is going, not where it has been.”
Senior Associate Dean for Medical Education Carla Lupi, MD, and Assistant Dean for Medical Education Deepu Gowda, MD, MPH, framed the retreat as part of an ongoing series of curriculum gatherings that build KPSOM’s community of practice. Good curriculum and integration, Lupi emphasized, “requires community. It does not happen any other way.”
