An intentional dual track curriculum model proposal to address technical readiness and concerns in healthcare AI education

The future of medicine is being written in code. Artificial intelligence is already helping doctors read X-rays, predict which patients are at risk of falling ill, and suggest the best treatments. As these tools become common in hospitals, a critical question arises: how do we teach the next generation of doctors to use them? For decades, experts have relied on a simple idea to explain how people accept new technology: the more you know about it, the less you fear it. This logic suggests that if we simply show students how AI works, they will become comfortable with it, their worries will fade, and they will eagerly adopt it into their practice. It is a straightforward path from ignorance to acceptance.

However, the reality of training doctors may be far more complex. Medical schools are now scrambling to add artificial intelligence to their curricula, but they are doing so based on the assumption that familiarity solves everything. If a student learns the basics of how an algorithm makes a decision, the theory goes, they will trust the machine and stop worrying about the risks. This belief drives the design of many new courses, which focus heavily on technical literacy. But what if this assumption is wrong? What if learning about AI does not make the deep, professional worries about patient safety and fairness go away? Understanding this distinction is vital, because if we teach doctors only how to use the tools without addressing their legitimate fears, we risk creating a generation of clinicians who are technically skilled but ethically unprepared.

Researchers at the University of Miami decided to test this assumption directly. They gathered a group of 389 first-year medical students, just as they were about to attend a workshop on artificial intelligence in healthcare. Before the students learned anything new in the session, the researchers asked them a series of questions. They wanted to know if these students had ever taken a class or attended a seminar about AI before. They asked how familiar the students felt with the technology. They also asked about the students' attitudes toward the future of AI, how prepared they felt to use it, and, crucially, what specific concerns they held. The researchers looked for six major worries: the fear that AI would make patient care feel cold and impersonal, the risk of unfair bias in the algorithms, the security of private patient data, the danger of doctors becoming too dependent on the machines, the protection of patient privacy, and the general lack of trust between patients and providers regarding these tools.

The results revealed a split in the students' minds that challenges the standard way we think about learning. On one side, the researchers found that students who had prior exposure to AI, such as a previous class or workshop, did report feeling more familiar with the technology. This familiarity, in turn, was strongly linked to positive feelings. Students who felt they understood AI were more optimistic about its future, felt more prepared to use it in their careers, and were more likely to say they intended to use it. In this sense, the old idea holds true: knowing more about the tool makes you more ready to use it. The depth of their understanding mattered more than just having attended an event; it was the feeling of being familiar with the concepts that drove their confidence.

But on the other side of the equation, a different story emerged regarding their fears. The study found that concerns about AI were not rare; they were nearly universal. Between 53% and 82% of the students expressed worry about specific issues, with the fear of losing the human touch in patient care being the most common, cited by 81.5% of the group. The researchers expected that as students became more familiar with AI, these worries would shrink. They expected that knowledge would act as a cure for anxiety. Instead, they found that the worries did not go away. Whether a student knew very little about AI or felt extremely familiar with it, their level of concern remained exactly the same. A student who felt very confident in their technical knowledge was just as worried about algorithmic bias and data privacy as a student who knew very little.

This finding suggests that the path to becoming a doctor who uses AI is not a single road. The researchers propose that there are two separate tracks that run parallel to each other. The first track is about building readiness. This is the path where education helps students understand how the technology works, how it fits into a hospital, and how to use it effectively. This track successfully builds confidence and the desire to adopt new tools. The second track is about managing concerns. This track deals with the heavy, serious questions of ethics, fairness, and the human relationship in medicine. The study shows that technical familiarity does not solve these problems. You cannot simply teach a student enough about code to make them stop worrying about whether an algorithm is treating a patient fairly or whether their data is safe. These are not gaps in knowledge that education can fill; they are professional judgments that persist regardless of how much you know.

The implications for how we teach medicine are significant. If schools continue to design AI courses with the idea that teaching the technology will automatically resolve all fears, they may be missing half the job. The researchers suggest a dual-track curriculum. One part of the education should focus on technical literacy, ensuring students know how to operate and understand the tools. The second part must be a dedicated, structured instruction on ethics, governance, and the human side of care. This second track needs to treat concerns not as obstacles to be overcome by more learning, but as legitimate professional obligations that must be addressed directly. The students in the study were able to be excited about the future of AI while simultaneously holding onto deep, reasoned concerns about its risks. They did not need to choose between being ready and being worried; they could be both.

This study does not claim that artificial intelligence is bad or that it should be avoided. It simply shows that the way we prepare doctors for it needs to change. We cannot assume that a student who is good at using a machine is also good at navigating the ethical dilemmas that machine creates. The research suggests that true readiness in the age of AI requires two distinct kinds of learning. One builds the skill to use the tool, and the other builds the wisdom to question it. By separating these goals, medical schools can ensure that their graduates are not just technically proficient, but also ethically grounded, capable of embracing the power of artificial intelligence without losing sight of the human beings they are there to serve.