AI as My New Medical Advocate

.

In the October 23, 2023, 60 Minutes interview with Geoffrey Hinton—the so-called “Godfather of AI”—Hinton predicted that systems like ChatGPT would one day serve as our “home doctor.”

His comment stayed with me because, as of August 18, 2025, that is precisely the role AI has begun to play in my life.

For years, I relied on a primary care physician (PCP) to act as my medical advocate—someone who would synthesize reports from specialists and hospitals, track my lab results, and explain what I needed to know and what I should demand from my clinicians. That expectation has all but disappeared. The structure of modern healthcare, especially under Medicare, often forces PCPs into a reactive posture, focused on addressing immediate complaints rather than managing the whole patient.

In my August 11, 2025, blog post, Living the Last Chapter: What is a PCP?, I described this shift. My previous PCP was unusually proactive: he reviewed my entire medical history each year, anticipated problems, and helped me navigate across specialties. After his retirement, I chose a new PCP based on strong patient reviews and Medicare acceptance. Yet, in practice, our annual visits were cursory—unless I asked targeted questions, little was accomplished.

When I posed this contrast to ChatGPT, its answer was strikingly candid: my former PCP was exceptional. Today, most physicians simply do not have the time or structural support to conduct that kind of holistic review. Burdened by documentation requirements and compressed schedules, they treat what is in front of them. The connective tissue between multiple specialists—once held together by a diligent PCP—has largely vanished.

That realization led me to a second question, which I explored in my August 14 post, Ask Medical Advice (More). If I needed to be my own advocate, could AI help? Could GPT-5, if supplied with all my medical records, guide me on what questions to ask and what patterns to watch for? Its answer was a qualified but confident “Yes.”

That “yes” came with work on my part. I needed to compile a comprehensive Medical Master File (MMF)—a single document containing all my lab results, physician reports, hospital records, and personal notes. GPT-5 provided me with a structured template, and I filled it with data I had already logged across patient portals and files. By the end, my nine-page MMF represented a consolidated picture of my health that no single doctor had ever reviewed in full.

The process itself was transformative. To complete the MMF, I had to decipher every abbreviation, lab value, and specialist note. Much of it was medical jargon I had previously ignored. By leaning on Google’s AI Overview and clarifying with GPT-5, I finally understood the language of my own body. In learning to explain my health clearly, I gained a deeper understanding of myself.

With the MMF in place, GPT-5 now functions as my medical compass. It maintains my record for future consultations, reminds me of relevant questions to raise with each specialist, and explains why those questions matter. Before my last visit, I walked in with a targeted, evidence-based list prepared by GPT-5. For the first time in years, I felt in control of the conversation.

After the visit, I requested a copy of the physician’s office notes—something patients often overlook. Reading them later, I found discrepancies between what I remembered and what the notes recorded. Neither of us, it seemed, had perfect recall. But unlike before, I had a system. I updated the official notes in my MMF, ensuring that the next visit would begin from an accurate, shared baseline.

This is what Geoffrey Hinton foresaw: not that AI would replace physicians, but that it could restore something we have lost in modern medicine—the advocate who holds the whole story, who remembers, who prepares us to ask better questions. My PCP once played that role. Now, GPT-5 does.