I'm a cardiologist, and here is the case for AI concierge medicine. It starts with an essay everyone in American healthcare should read this week: "Want Concierge Medicine? AI Can Deliver It," by Tom Rosenblatt in @WSJopinion. He's the founder of the health-longevity company Being and sits on a federal advisory panel on tech-enabled care — and he named the problem I've watched break good doctors for 25 years. h/t Eric Topol. The case, point by point. Attentive care has quietly become a luxury. When Rosenblatt went looking for a primary-care doctor in New York, the physicians with open calendars had turned to concierge practices — $12,000 a year at one, $24,000 at another. That is the going rate now for what used to just be called having a doctor: someone who knows you, sees you promptly, and has time to think about your results. More than 100 million Americans don't have that at all. This isn't a villain story. It's broken math. Primary care pays less than almost any specialty. Most medical students graduate owing over $200,000. So fewer choose the field, urgent care mops up the crises, and no one is paid to prevent anything. Here's the number that should end the debate: to deliver all the recommended preventive care for a normal patient roster, a primary-care doctor would need more than 14 hours a day. Every day. It's not that we're failing at prevention. It's mathematically impossible as built. And the country is paying for it in bodies. Only 7% of American adults have healthy blood pressure, blood sugar, cholesterol, and weight — all four. The other 93% are sitting on a slow slope from early risk to real disease. And the cruel irony is we know what pulls people back: move most days, eat mostly unprocessed food, sleep enough, stay connected. For early prediabetes and mild hypertension, those four habits can rival a prescription. But no one has 20 unhurried minutes to tailor that advice to you anymore. AI changes the arithmetic — because it doesn't get tired at hour 14. A well-built system can read years of your labs, imaging, and wearable data in seconds and catch the upward drift in five numbers at once, long before any single one crosses a red line. The heart attack that shows up at 55 was often visible in the data at 40. We simply never had eyes patient enough to watch continuously. One in three American adults already asked an AI for health advice last year. But I'm a physician, so I'll say the hard part out loud. On real clinical cases, some AI models have given safer advice than generalist doctors. And other models made a serious mistake in more than 1 in 5 cases. Right now most health AI dodges federal review entirely by calling itself a "wellness product." That gap — between the good models and the dangerous ones — is the whole ballgame. A tool this powerful cannot be sold on trust alone. Which is why the case isn't "download an app." It's build the guardrails. Rosenblatt lays out three, and as a doctor I'd sign my name to all three. Records that follow the patient automatically, so you own your full history and decide who sees it — human or AI. A public crash test for medical AI: rotating real clinical cases, scored and published, the way we rate cars for safety, so you know which system to trust with your life. And payment for outcomes instead of visits — Medicare's new ACCESS model went live July 5th, paying only when your blood pressure actually drops and your A1c actually comes down. That's the missing piece that turns an AI's warning into care someone is finally paid to deliver. Put it together and here is the whole case in one line: A doctor's time is scarce and expensive. An AI's time is nearly free. The continuous, prevention-first, always-watching attention that today costs the price of a luxury car could become available to anyone holding a phone. This is not AI replacing your physician. I have spent my life at the bedside; nothing replaces the human hand on a frightened chest. This is AI giving every person the second set of eyes and the personal foresight that, until now, only the wealthy could buy. The technology is in our pockets. The data can be made portable. The accuracy can be proven and publicly scored. The incentives are starting to change. The pieces exist. So the future of medicine gets to stop being a privilege. It becomes a right. That's the case. And it's the reason I'm building what I'm building — because I don't want to be brilliant at the worst moment of your life. I want to make sure that moment never comes. Thoughts? @rabois