Physician and enterprise product leader. Author of the Agentic AI for Product Leaders series. Writes on AI, healthcare, and what changes when software starts acting. 15+ years at SAP, Walmart, and regulated health tech.
I’m seeing a flood of self-proclaimed “AI experts” with little real depth. After decades building enterprise systems, I still hesitate to claim the title. In healthcare AI, experience matters more than hype. Projects fail from old debts and governance gaps, not lack of buzzwords.
Digital twins are moving from factories to hospitals. We’re building replicas of patients, modeling disease in real time, and of physicians, capturing decades of expertise. The twin extends capacity while preserving oversight. What will happen when they no longer need us in between?
The narrative says AI pilots fail in healthcare because the industry is slow and resistant. It is backwards. No industry changes faster than healthcare when the change actually works. The problem is not resistance. It is that most AI pilots have not earned the right to change it.
We can’t talk seriously about AI in healthcare without understanding the data infrastructure behind it. From HL7 v2 to FHIR, SNOMED, and MCP, I break down how clinical data actually moves and why every healthcare leader needs to grasp this before approving the next AI pilot.
The bottleneck in tech is no longer writing code, it’s deciding what to build. As AI collapses development time, value shifts upstream to product judgment. In healthcare especially, models work, but projects fail without real product thinking. The new question isn’t can we build it, but should we.
A 57-year-old liver transplant patient arrives confused and lethargic, with asterixis and ascites. No experienced clinician makes a diagnosis from one signal, a single heart rate, a single lab, a single image. But much of healthcare AI is still built that way. The clinical reality is multimodal.
At 2am, every product manager runs the same simulation, an imaginary room of stakeholders. The problem? You built the room. AI can become the same echo chamber. There are five levels of working with AI, and only the higher ones truly challenge your thinking.
In healthcare AI we focus on algorithms, models, accuracy. The quieter question: whose data can the AI actually see? The answer has less to do with the model and more to do with how you receive care. Your insurance type may determine whether your care is AI-ready at all.
I ran a full design thinking workshop on a Sunday afternoon, no flights, no sticky notes, just nine AI agents configured to disagree. For under $10, they reframed the problem entirely.
AI isn’t just augmenting expertise, it may be eroding it. From physicians to developers, evidence shows short-term gains can come at the cost of long-term competence. The real risk isn’t dependency, it’s losing the ability to function without AI.
AI won’t replace physicians, but it will redefine them. As knowledge shifts from memory to machine, the xPhysician emerges: an AI-augmented generalist who combines deep expertise with end-to-end ownership, reshaping how medicine is practiced and learned.
The FDA’s January 2026 CDS guidance isn’t deregulation, it’s smart, risk-based clarity. AI tools that support physician judgment without black-box automation can move faster. High-risk diagnostic systems remain regulated. That’s not stepping back, it’s focusing oversight where it truly matters.