Exploring how physicians can make themselves ‘AI-proof’

Over the past couple of years, artificial intelligence (AI) has become a big buzzword throughout all industries in America, and healthcare is no exception. As a 2026 DO graduate, I have gotten the opportunity to witness the integration of AI into clinical practice, including ambient AI scribes used in clinics to generate patient notes, software that helps detect polyps during colonoscopies and the use of OpenEvidence for clinical decision support.

A popular form of AI is the large language model, which is essentially a highly sophisticated word prediction system trained on vast amounts of data. These models are constantly iterated and optimized to produce accurate and useful outputs, making them excellent at recognizing patterns and following algorithms.  

As a future clinician, this is particularly relevant when we consider that a notable portion of a physician’s workflow operates algorithmically, more so in some specialties compared to others; however, every single specialty still possesses an algorithmic nature on some level.

Hospitalists follow diagnostic and management algorithms for patients with sepsis, electrolyte disturbances and blood clots. Emergency physicians run mental decision trees for patients presenting with chest pain, headache and lightheadedness. Even surgeons follow a highly systematic, step-by-step process when performing procedures like removing a patient’s appendix, moving layer-by-layer in a methodical and predictable way. AI excels at supporting physicians who are doing this sort of algorithmic work.

Now just to be clear, this is not to say that everything a physician does is purely algorithmic or can be easily translated to AI. That is far from the reality, as factors like human nuance, ethical judgement and the sheer complexity of individual patient contexts remain extremely difficult to train AI on.

At the same time, most of the profession has come around to the idea that AI has the potential to enhance/is already enhancing certain aspects of the clinical workflow.

In this article, I wanted to share insights from the many recent conversations I’ve had with medical students and practicing physicians about AI in medicine. Along with my experiences with observing AI use in the clinical setting, these conversations have informed the thoughts I share below about how the osteopathic medical profession can thrive and help lead medicine in the age of AI.

Essentially, by doubling down on the empathetic, patient-centered care that DOs are known for while learning and exploring how AI can help lighten our administrative burden, DOs can excel in this emerging environment.

An uncertain future

When I’ve discussed AI in medicine with my former medical student classmates and physician mentors, many express anxiety, excitement and uncertainty.

Some cite concerns about the potential of AI to put them out of a job: If AI is good at algorithmic work and a lot of the work physicians do is algorithmic in nature, what does that mean for my role as a physician? Will physicians lose professional autonomy to AI? Even scarier, could AI potentially replace the need for human physicians altogether? More importantly, will the increasing the integration of AI into clinical care compromise patient safety or decrease quality of care?

These fears are valid. And they are questions that deserve appropriate answers from our leadership and which will require multidisciplinary collaboration among physicians, machine learning and artificial intelligence experts, healthcare administration, patients and beyond.

With any advancement in technology or change, people get uncomfortable. People don’t like change. But as the record of human history shows, the only constant in this world is change. People who are mindful of their ego and make an effort to see the potential of AI objectively will be the ones who will thrive in an increasingly AI-integrated healthcare system.

Instead of asking “Is AI going to make physicians obsolete?” the better question is “How can I adapt and prepare for a future that includes AI?” Rather than feeling helpless and being in denial about the inevitable, the most productive thing that future and current physicians can do is approach AI not with fear, but with curiosity and intentionality. The bottom line is that AI is here and it is here to stay whether we like it or not. I am of the opinion that as AI advances, the role of human physicians will become more important, not less.

Embracing an ‘AI-proof’ mindset

But the only way I see this happening is for future and current physicians to commit to  becoming “AI-proof.” When I say “become AI-proof,” this means developing the skill sets that will remain relevant to the practice of good medicine no matter how advanced AI gets. Such skills include but are not limited to the humanistic “soft skills” centered on human connection, empathy and reassurance.

Here are a few hypothetical future patient scenarios that reiterate my belief in such AI-proof competencies:

• Scenario A: A 26-year-old schizophrenic patient is admitted to the psych ward in a combative and noncooperative state. An AI interface will fail compared to a good human psychiatrist, who will fare better in conveying calmness and understanding. 
• Scenario B: A 58-year-old mother of two, newly diagnosed with breast cancer, is considering her options. Sure, AI can present her with accurate medical information, as well as outline risks, benefits and her alternative options. But the patient is overwhelmed, distressed and scared. She is in analysis paralysis and her heart has been heavy since the news of her diagnosis months ago. In this patient context, the role of the human physician as extended family becomes extremely important and necessary. That reassurance and empathy is what this patient needs, and only a human physician can give that to this patient.
• Scenario C: A 41-year-old woman has been having trouble sticking to her exercise routine since her divorce a year ago. She already logically knows what next practical steps she needs to make, but she feels disconnected from her underlying motivation and purpose that would allow her to start taking those next steps. A human physician skilled in motivational interviewing would be indispensable for helping this patient reconnect with that motivation and her values in a way that no algorithm can replicate. 

The above scenarios illustrate an important point about how skills like bedside manner and the humanistic element will become far more important in the future as AI gets more advanced. Because at the end of the day, humans will still be humans. People will still need people. And that is why the role of human physicians for human patients can never be fully taken away by AI.

In addition to building these human skills, future and current physicians should also develop basic AI competencies, including becoming educated on where they can leverage AI into their daily workflow as well as learning the limitations of AI.

While individual effort on the part of future and current physicians is extremely beneficial, more than this will be needed to truly prepare America’s physicians to thrive in the age of AI. Change will be required on a systems level. AI education and training will need to be integrated into medical schools, residency programs, accreditation bodies and healthcare systems.

Looking ahead

Learning AI competencies needs to be established as a prerequisite learning milestone for advancement, just like other core clinical skills like collecting an appropriate history, performing a physical examination and presenting on a patient. I also believe that an AI literacy course that teaches students the pros and limitations of AI needs to be formally integrated into medical school curriculum as its own course, alongside foundational courses in pharmacology, physiology and anatomy.

Whether we like it or not, AI will reshape the way healthcare is delivered in the next few decades. The ones who embrace it and who build their AI proof humanistic skills and people skills will be best prepared to succeed. Leadership in medical schools, residency programs, accreditation bodies and medical licensing boards need to take the initiative in integrating AI competencies into training, so that human physicians are prepared to navigate our rapidly changing, AI-driven landscape of healthcare.

Related reading:

OsteopathicAI: How a professional standard for AI can strengthen our commitment to whole-person care

15 minutes under paper weight: Why medicine’s most meaningful moments can’t be charted