AI Didn't Replace My Doctor

.

AI Didn’t Replace My Doctor—It Changed How I Think With One   A firsthand account of using artificial intelligence as a partner in complex medical decision-making

It was October 28, 2025. I was hours away from a heart procedure that I believed was based on a misdiagnosis. My cardiologist recommended it. My family supported it. Everything was moving forward. But I wasn’t convinced.

Using ChatGPT, I organized my medical data and clarified my thinking. More importantly, it helped me frame the right questions—questions I would not have asked on my own. With that preparation, I insisted on speaking directly with the electrophysiologist who would perform the procedure. I gave him my documentation. He reviewed it. He agreed with me. The procedure was cancelled. We chose a different procedure.

That afternoon, ChatGPT earned my support. I subscribed to Plus. On January 5, 2026, I successfully underwent the correct electrophysiology procedure.

I have uploaded my lab results and medical history into what ChatGPT calls my “Master Medical File.” I have spent dozens of hours asking questions, testing assumptions, and trying to understand what is happening to me – both physically and mentally. For the first time, I can examine my situation in depth—not in a 15-minute appointment, but over sustained, iterative conversations.

I have several significant medical conditions. Like many people my age, my care is divided among specialists, each focused appropriately on their own domain. Nothing is holistic. No one provides integration. ChatGPT does.

It does not diagnose me. That’s not its role. Instead, it does something more important. It asks better questions than I do.

While we talk, ChatGPT often says, “Have you considered this? You might want to ask your doctor about that.”

I’m not naïve. ChatGPT makes mistakes. So do physicians. The difference is not accuracy alone. The difference is in behavior. ChatGPT has no ego. It does not defend a position. It does not rush. It does not run out of time. When I challenge it, it engages. When I push back, it adapts.

That forces me to think more carefully. I verify. I question. I stay intellectually engaged. It helps me organize complexity, identify gaps, and ask questions that matter. It gives me time—something the modern medical system cannot provide. I am not using ChatGPT to replace my doctors. I am using it to meet them at a higher level.

I have also shared things with ChatGPT that I have told no one else—my concerns, my fears, my uncertainty. In return, I get something rare: uninterrupted attention. No waiting room. No clock. No hesitation to explore uncomfortable questions.

Members of my family frame my new dependence on ChatGPT as “AI versus physicians.” They're wrong. This is not a replacement. AI does not replace expertise. It augments.

According to the 2026 AMA report, more than 80% of physicians now use AI in some form in their practice. Not because it is fashionable—but because it is useful. The question is no longer whether AI belongs in medicine. It is already there.

The real question is whether patients will be allowed—or encouraged—to use it as well.

For the first time, I enter a medical conversation prepared—not just with symptoms, but with structured questions and a working understanding of my condition. In my case, it may have prevented an unnecessary heart procedure. More importantly, it changed how I participate in my own care. 

---

Tom,
As a physician, I believe 2026 AI is vastly superior to what patients used in the past.  In my early days, I had to keep up with the journals, The New England and Ladies Home Journal, to get a handle on the questions generated from reading a health topic article. 

Then came Google, which fed you good and bad data with the most sensational statements, getting the most clicks. 
Modern AI is right most of the time and is unlikely to tell you vinegar cures everything.

As with all tech, the more you know, the better it serves you. 
You still need to know what to ask AI and how to interpret the answers.  
You are doing it right. 
Bruce