Category: AI
Essays on what AI makes possible in work and everyday life. Explore changing jobs, building software, and the role of human judgment when machines can generate more than we can verify.
Selected reading
- The Jagged Frontier of Medical AI — Why medical AI progresses unevenly across different kinds of work.
- AI Jevons Paradox: Why AI May Create More Work, Not Less — What programmers and medical scribes reveal about changing jobs.
- The Blinking Cursor Returns: Finding My Childhood in the AI Agent Era — Rediscovering the pleasure of making things with AI agents.
- Your Gut Is a Value Function — How experience shapes the judgment we call intuition.
-
The Jagged Frontier of Medical AI
·
AI can reason faster than medicine can prove an answer right. A framework for understanding where medical AI can act independently, where it needs supervision, and why biology keeps its own clock. Read more
-
An Audience of One
·
AI makes it possible to build the tool no company would make for you. On personal software, the judgment only its maker can supply, and why an audience of one is worth building for. Read more
-
The Patient Will See You Now
·
Patients have spent years becoming better informed. AI is beginning to help some of them make the first move in their own care. What changes when a patient arrives with a draft plan, and who gets that chance? Read more
-
AI Jevons Paradox: Why AI May Create More Work, Not Less
·
When AI makes a task cheaper, the work around it may grow. What software developers and medical scribes reveal about shrinking roles, expanding responsibilities, and the conditions that give human work a lasting place. Read more
-
The Blinking Cursor Returns: Finding My Childhood in the AI Agent Era
·
A blinking cursor has brought back the feeling of my first computer. Building with AI agents reminds me why I loved the command line, and raises a harder question about bringing that freedom into healthcare. Read more
-
The Humbling Math of Health AI: Why ChatGPT Can’t Grade Your Heart Yet
·
Years of Apple Watch data can look like a complete picture of your health. Making sense of it is another matter. What the work behind narrowly validated health features can teach us about the limits of a fluent AI answer. Read more
-
When You Upload Your Medical Records to AI, Who’s Actually Protecting Them?
·
A privacy promise is only one part of protecting a health conversation. Drawing on my experience at Apple and Verily, I explore what changes when medical records move into consumer AI, and why the architecture matters. Read more
-
A Year of Talking to Computers
·
I shipped four iOS apps in a month without being an engineer. A year of AI coding taught me the bottleneck moved from building to judgment. Read more
-
Your Gut Is a Value Function
Your gut isn’t mysticism—it’s a value function trained on real outcomes. Why AI decision making can’t replicate human intuition built over decades. Read more
-
Strategy in the Age of Infinite Slop
AI generates infinite strategy options but can’t judge which survive reality. Why AI slop won’t replace human judgment in content strategy. Read more