window.dataLayer = window.dataLayer || []; function gtag(){dataLayer.push(arguments);} gtag('js', new Date()); gtag('config', 'UA-97641742-42');

LMS Essay Contest 1st Place, Medical Student Category

On November 17, 2024, the New York Times declared, “A.I. Chatbots Defeated Doctors at Diagnosing Illness.” The hallowed fields once regarded as impervious to automation—medicine, art, education, law—appeared demystified, their treasured knowledge and traditions becoming mere computer code. Given patient histories alone, AI reached the correct diagnosis with 90% accuracy. The doctors scored 74% on their own and only reached 76% when allowed to seek help from their AI competitor. Has the prophesied era of “Dr. Robot, MD” crept closer, or has it already arrived?

We have long feared a world where cold, unfeeling robots impose their algorithms on humans without care for individuality or context. Adaptive thinking was our prized trait, elevating humanity above animals and distinguishing us from machines. Perhaps we are the ones applying standards irrespective of nuance. Maybe that is why we feared robots would do the same. We have been stuck in our ways since AI’s inception in the 1950s. We were too busy contemplating whether it would help or harm medicine to notice how machines began to win the competition.

As a medical student, I see the clinical potential of AI firsthand. I use AI daily to help me with various tasks, from creating study guides and practice questions to generating promotional materials for fundraising events. Each interaction teaches me something profound: AI is not a replacement for medical expertise but rather an amplifier. We can already see how the notion of patients coming into the clinic with diagnoses from ChatGPT is transitioning from an alleged inevitability to our current reality. Looking back at the rise of the Internet and WebMD, the evolving role of AI in medicine mirrors how we responded to other technological advancements—from doubt to acceptance to integration. After all, are we still debating the fundamental value of the computer in clinical practice? Or the Internet?

This decades-old quandary has distracted us while AI starts to eclipse our diagnostic approaches. Can AI enhance medical care, eventually improving our test scores to match it? Indeed, we could certainly focus on bridging the gap to 90%, enhancing ourselves to match the skills we currently lack. There is promise in the idea of equipping doctors with superior diagnostic tools and appropriate training.

The narrative of AI versus doctors, however, might be a misleading one: more appropriate for driving newspaper engagement than informing the philosophy of medicine. While we fall prey to the debate of whether artificial intelligence will enhance or deter medical care, perhaps we keep asking the wrong questions. Have we pondered how medical care might enhance or deter AI? Might our patients be better served by focusing on that missing 10% AI could not achieve on the test? Could we help create a world exceeding what either of us could accomplish alone?

Instead of viewing AI as a competitor that scored 90% against our 74%, we should focus on how combining our capabilities could achieve 95% or higher. The missing 10% in AI’s performance represents the human elements of medicine—the intuition built from years of patient interaction, the ability to read between the lines of a patient’s history, and the emotional intelligence that no algorithm has yet mastered.

Just as we train medical students and residents, we must approach AI as a “colleague” in training, one that requires supervision, guidance, and—most importantly—our clinical wisdom. We do not expect resident physicians (or even the attending) to work in isolation. Offering guidance, sharing experience, and refining clinical judgment are essential to the formation of a skilled medical practitioner. To an even greater extent, collaborative education is inherent to the professional development of medical students. Likewise, AI requires our active participation—not tacit, begrudging, or hostile acceptance—to reach its full potential in healthcare. When we treat AI as a trainee rather than a replacement, we open possibilities for unprecedented advances in patient care.

The Times article revealed a telling detail: doctors only improved their diagnostic accuracy by 2% when using AI assistance. This modest improvement shows we have yet to learn how to effectively collaborate with AI’s burgeoning capacity for clinical reasoning. As doctors do not currently understand AI, the converse is necessarily true given that these computer models are trained on human input.

Our role as physicians must evolve. Beyond treating patients, we now have the responsibility to ensure that AI systems are trained with accurate clinical information and ethical guidelines. Medical students, residents, and attending physicians must become fluent in both medicine and AI collaboration, understanding how to leverage these tools while safeguarding the fundamental human nature of medicine.

This perspective transforms the narrative from competition to collaboration. Instead of asking whether AI will enhance or deter medical care, we should focus on building a future where human expertise guides AI development and AI augments human capabilities. The goal is not to match AI’s current performance but to create a partnership that exceeds what either humans or machines could achieve alone. Perhaps then, the era of “Dr. Robot, MD” will give way to something far more powerful: an era where humans and AI are partners in care.