I have taught AI to more than 5,000 medical and dental students, physicians, and other healthcare professionals across the United States. I have run workshops, webinars, and lectures at hospitals, universities, conferences, and online. I have spoken to first-year medical students who have never seen a patient and to senior physicians with 30 years of clinical experience. I have taught in rooms full of enthusiasm and rooms full of scepticism.
After all of that, I want to share what I have learned — not about AI, but about how medical professionals think about AI, where the resistance comes from, and what actually changes minds.
The First Reaction Is Almost Always Scepticism
When I walk into a room of medical professionals and introduce AI, the most common first reaction is polite scepticism. Senior physicians have seen many technologies promised as transformative and delivered as disruptive without being genuinely useful. Junior doctors are enthusiastic about technology in principle but uncertain about what it means for their careers. Medical students are the most open, but they often have the least context for evaluating AI claims critically.
The scepticism is healthy. The problem is that it often stops at 'I don't trust AI' without engaging with the more important questions: which AI applications are trustworthy and why, where AI has demonstrated genuine clinical value, and what the cost of not engaging with AI literacy is for a physician's career and their patients.
What Medical Professionals Fear Most
Two fears consistently emerge. The first is displacement — the worry that AI will make physicians less necessary. The data does not support physician replacement by AI in clinical practice. It supports physician augmentation. The second fear is accountability — who is responsible when an AI recommendation is wrong?
This second concern deserves a clear answer: the accountability framework in medicine has not changed. The physician who makes the clinical decision is accountable for it. AI tools are decision support, not decision authority. Understanding this distinction is fundamental to using AI safely — and it is something I teach explicitly in every session.
The Moment It Clicks
In every workshop I run, there is a moment when the room shifts. It is usually when I show a specific, concrete example of AI doing something that visibly saves time or catches something a physician missed — a sepsis AI flagging patient deterioration hours before clinical signs became obvious, or a student interacting with an AI case study that responds dynamically to their clinical reasoning.
"When the abstract becomes concrete, the conversation changes. The question stops being 'should I trust AI?' and starts being 'how do I learn to use it well?'"
— Dr. Chandra Sekhar BondugulaWhat the Curriculum Needs to Include
Based on 5,000+ hours of teaching AI to medical professionals, here is what the curriculum must cover:
- How to evaluate an AI tool critically — the evidence base, validation studies, and known failure modes
- Decision support vs decision authority — this determines accountability and appropriate trust levels
- What AI cannot do — it cannot examine a patient, read a room, or build a therapeutic relationship
- How to document AI-assisted decisions — increasingly important for patient safety and medicolegal protection
- The economics of AI in healthcare — physicians who understand this are better advocates for their patients
What I Would Tell Every Medical Student Today
If I could give every medical student one piece of advice about AI: engage with it early and engage with it critically. Do not wait until you are in clinical practice to develop AI literacy. The doctors who will matter most in the next decade are not the ones who know the most facts — an AI can retrieve facts faster than any human. They are the ones who combine strong clinical fundamentals with the wisdom to use AI well: knowing when to trust it, when to question it, and how to explain both decisions to their patients.