Logbook.
Notes from the work. Essays on AI experience design, agentic systems, and what comes after the screen.
01
What AI hardware should feel like
A defense of treating AI devices the way we treat the things we already keep around us — wedding rings, reading lamps, favorite coats. Notes from building Fountain on why the next decade of computing will be defined by what AI looks like off-screen.
Draft
02
Making AI feel like part of the work
When AI starts doing the work for you, the question stops being how do we talk to it and becomes where does it appear, and what does it leave behind. Lessons from designing the agentic experience layer at enterprise scale.
Draft
03
Trust is a design problem
Most of what we call AI safety is actually an interaction design question — about consent, reversibility, and what an agent is allowed to do without asking. A pattern language for the line between assistance and autonomy.
Draft
04
The generative canvas
A UI primitive for the agentic age — what it is, why it matters, and the tradeoffs we're still working through. Notes from shipping it.
Draft