David Toro-Tobon, MD, recalls the moment he was at the crossroads of artificial intelligence and medicine. When he arrived at Mayo Clinic in Rochester, Minnesota, for his endocrinology fellowship in 2022, AI was no longer a remote concept discussed in engineering circles or speculative headlines. Suddenly, it was everywhere—embedded in conversations about patient care, research, workflow efficiency, and the future of clinical decision-making. “It started a little bit of an internal revolution,” he says.

Today, Dr. Toro-Tobon is an assistant professor of endocrinology at Mayo Clinic, where he serves as an investigator at the Care and AI Laboratory. But his journey began thousands of miles away in Colombia, where he attended medical school at CES University in the city of Medellín, then moved to the US for an internal medicine residency at Georgetown University in Washington, DC. From there, he pursued endocrinology at Mayo Clinic, eventually developing a clinical and research focus in thyroid disease, and where he currently serves as an assistant professor of endocrinology. 

What drew him toward AI was not technical expertise, but curiosity—and eventually, concern.

“My knowledge of AI 5 years ago was minimal,” he says. “I’ve always been a person who is kind of literate in terms of using regular computers and smartphones, and I always wanted to keep up with the latest trends. But other than that, my knowledge in terms of AI was limited.”

Still, Dr. Toro-Tobon quickly sensed that health care was entering a transformative period. Generative AI tools were becoming more accessible, data-driven medicine was accelerating, and institutions across the country were beginning to invest heavily in AI initiatives.

Pivotal Moment

What surprised him most was the growing disconnect between two groups that would ultimately need one another: data scientists building health care AI systems, and clinicians expected to use them.

“We have a lot of data scientists that have pretty much non-existent clinical knowledge trying to do AI for health care,” he says. “And on the other extreme, we have a lot of clinicians who are expert clinicians with very limited AI knowledge trying to get involved in AI. And these two worlds don’t communicate easily.”

That realization became a defining moment in his career. During early AI training experiences through courses at MIT and Harvard, Dr. Toro-Tobon found himself in discussions with data scientists reviewing AI-related medical studies published in top-tier clinical journals. The reaction from the technical experts astounded him.

“They would pretty much trash these papers and say these would have never been published in a data science journal,” he says.

For a young physician immersed in evidence-based medicine, the experience was jaw-dropping. Here were studies celebrated within medicine as groundbreaking innovations yet viewed by technical experts as methodologically weak or clinically overhyped.

“To me, this was a huge eye-opener,” Dr. Toro-Tobon says. “I thought, ‘We have a big problem here.'"

That problem, he believes, goes beyond research methodology. In health care, especially across the wide field of endocrinology, AI cannot simply be “good enough.” The stakes are fundamentally different.

“The tolerance for error here is not the same,” Dr. Toro-Tobon says. “We cannot have margins of error as relaxed as you can have in other industries.”

This philosophy now shapes much of his work. Rather than approaching AI as a technological novelty, he views it through the lens of patient safety, scientific rigor, and clinical usefulness. His goal is not simply to promote innovation, but to ensure that innovation is trustworthy.

Mayo Clinic recognized the increasing need for physician leaders capable of navigating both medicine and data science, and Dr. Toro-Tobon became part of a new generation of clinicians training at that intersection. He is currently completing a Master’s Degree in Artificial Intelligence in Health Care at the Mayo Clinic Graduate School of Biomedical Sciences.

The academic training has transformed his research portfolio. Much of Dr. Toro-Tobon's work at the Care and AI laboratory now incorporates AI-based methodologies, including natural language processing and computer vision applications. His overarching goal is to tackle the growing clinical challenge of thyroid cancer overdiagnosis and overtreatment, ensuring that AI interventions translate directly to safer, more personalized patient care, minimizing unnecessary procedures for low-risk cases while optimizing treatment and outcomes for patients who require definitive intervention.

Yet despite the technical sophistication of his research, Dr. Toro-Tobon remains focused on practical clinical impact. He is especially enthusiastic about predictive modeling and the possibility that AI could help endocrinologists answer questions that have historically required years—or decades—of prospective study.

“Traditional prospective studies remain the gold standard, but they are costly, time-consuming, and often impractical,” he says. “We have a technology now with the power to look at massive amounts of data going back 30 years, and it’s possible you can get an answer that is very close to what would take a prospective study 10 years to achieve.”  

Dr. Toro-Tobon points out that many of the technologies making the biggest immediate difference in endocrinology have already quietly integrated into everyday clinical care. AI-assisted insulin pumps are improving glycemic management in real time. Generative AI documentation tools reduce administrative burdens and restore time for more patient interactions. “And that’s just the tip of the iceberg,” he says. 

Still, Dr. Toro-Tobon stresses the growing need for clinicians to understand AI’s limitations, biases, and appropriate applications. He worries that physicians may soon be inundated with AI literature and products without adequate preparation to evaluate them critically.

“How is it that we’re being prepared to understand what the new literature that is being thrown at us is showing?” he says. “How is it that we’re being prepared to use these tools in practice when they come?”

These questions have become central to his role on the AACE Endocrine AI editorial board, where he hopes to help shape the way AI-related endocrinology research is communicated to clinicians.

Part of that responsibility, he says, is identifying pseudoscience and ensuring that AI studies meet rigorous standards before reaching practicing physicians. But another equally important goal is accessibility.

“We need to make sure that what is coming through us is presented in a way that is understandable to a clinical audience,” he says. “What are the clinical implications? Is this ready for practice? Is this not ready for practice?”

Dr. Toro-Tobon’s enthusiasm extends beyond medicine and technology. He is an avid traveler—recent sojourns have included Peru and Sri Lanka—explores photography, and embraces the outdoors no matter the season, skiing in winter and biking in summer. 

Dr. Toro-Tobon says he has a broad vision for the future of endocrinology and AI. He is not advocating that physicians become software engineers, nor that technology replace clinical judgment. Instead, he sees AI as a powerful tool that, when carefully developed and responsibly implemented, can help clinicians deliver safer, more personalized, and more efficient care.

“It’s very easy to be tech savvy with AI,” he says. “So long as you know what you’re trying to accomplish and what you’re looking for.”