Care and computation are one subject

I was trained to take things apart. A doctoral degree teaches you to isolate one variable, hold everything else still, and report what moved. It is a good discipline, and it leaves a habit: you start to believe that the world is made of separable parts.

I work in two disciplines at once. One is the discipline of the laboratory: isolate, measure, report. The other is the discipline of the workshop: build the thing, put it in someone’s hands, watch what breaks. Neither is wrong. Neither is enough.

That is the whole argument of this site, so I may as well state it first. I do not think technology and care are two subjects that happen to meet in a few products. I think they are one subject, badly divided by the way we train people. Engineers learn to build without ever sitting with a person who is afraid. Physicians learn to sit with that person without ever being allowed to change the tools in the room. Both groups are then surprised when the tools are wrong.

I have spent the last few years on that seam. I co-founded a company for physicians’ careers, and I run a fleet of AI agents that does most of the engineering for my companies, which means I now write less code and more intent. I say these things out loud at conferences when asked, and I write them down here when not.

What follows is not a portfolio. It is the handful of ideas I keep returning to, written down once each. Read them in any order.