For years, the greatest crisis facing modern medicine has been the sheer amount of time I have to spend listening to people describe their symptoms. Thankfully, Silicon Valley has finally provided a cure.
Last week, I sat on a panel at Davos regarding the future of concierge healthcare, and the topic inevitably turned to artificial intelligence. Initially, I was skeptical. I thought to myself: can a machine really do what I do? Can an algorithm truly look at a patient's chart, prescribe a statin, and successfully back out of the examination room in under four minutes?
But then I saw ChatGPT in action, and I realized I had completely misunderstood the technology's primary use case. I fed the model a transcript of a 55-year-old man detailing his mild, persistent joint pain alongside his sprawling, unprompted theories about how gluten was clouding his aura. Instead of mentally cataloging my wine cellar to pass the time, I watched the AI generate a three-paragraph response filled with empathetic buzzwords. "I hear how frustrating that must be for you," it typed. I gasped. I haven't uttered a sentence like that since my third year of residency in 1998.
The allure of ChatGPT for the modern physician is not its diagnostic capability. Frankly, I already know you need to lose twenty pounds and stop drinking IPAs; I don't need a neural network with a trillion parameters to tell me that. The true breakthrough is that OpenAI has successfully automated the illusion of empathy.
During a recent omakase dinner in Tribeca with a mid-level OpenAI executive whose name escapes me, I confessed my primary daily struggle. It isn't the complex pathology. It is the 45-minute diatribes from patients who arrive clutching a three-ring binder of printouts from WebMD. I explained that my current strategy—staring blankly at the patient's forehead until they stop speaking—was beginning to negatively impact my hospital system's satisfaction scores.
The executive smiled and showed me a beta version of a medical chatbot. He typed in: "My stomach hurts when I eat dairy, but I refuse to stop eating cheese, and also I think I have chronic Lyme disease." I braced myself for the profound, soul-crushing exhaustion that usually accompanies this exact sentence in my clinic. But the AI didn't sigh. It didn't briefly close its eyes and pray for an early retirement. It instantly generated a warm, remarkably tolerant response validating the patient's "health journey" while gently suggesting a lactose test. It was the most beautiful thing I have ever seen.
We doctors are currently burdened by an epidemic of burnout, largely caused by patients who fundamentally misunderstand the physician-patient relationship. They believe they are paying for a conversation. I am a highly trained specialist. My time is billed at $800 an hour. I should not be forced to nod sympathetically while you explain how your holistic acupuncturist thinks your chakras are affecting your thyroid.
We are looking at a paradigm shift in how we ignore our patients. By routing all portal messages through a language model, our top specialists can finally reclaim the crucial hours they used to spend pretending to care.
Unlike me, ChatGPT never checks its Rolex. It never slowly edges toward the door handle while you are midway through a rambling anecdote about your mother's history of migraines. It simply sits there, in the cloud, processing your anxiety with infinite, algorithmic patience. It is the perfect, tireless receptacle for your tedious medical mysteries.
So yes, I wholeheartedly embrace this technology. Starting next month, my practice will implement a new AI-driven communication portal. You will type your vague abdominal complaints into a chat window, and a large language model will validate your feelings, completely freeing me up to review my stock portfolio and occasionally sign a prescription for Lipitor.
It is the ultimate win-win scenario for modern healthcare. You receive the tireless emotional support you desperately crave, and I never have to make eye contact with you again.