Things worth saying carefully.
Longer-form notes on design, teaching, AI, taste, and product work.
A prototype should answer one question
The fastest way to make a prototype useless is to ask it to prove the whole product. Pick the riskiest decision, make that part real, and let the rest stay rough.
The classroom is a product lab
Repeated confusion is product evidence. When the same explanation keeps coming up, the answer might be a tool, a framework, or an experience students can use without me.
People arrive mid-problem
No one enters a product as a blank user. They bring constraints, expectations, and a decision already in motion. Good interfaces begin there.
Taste without language is just a feeling
Most people know when something feels off. Fewer can name why. The gap between those two places is where designers actually get better.
AI is great at moving. It is not great at taste.
AI can help you ship faster, write faster, test faster. But knowing what to keep, what to cut, and what feels slightly off still takes judgment.
Legacy software doesn't need a total reinvention
It needs taste, restraint, and a better visual system. The hard part isn't making it look new — it's making it feel current without breaking the habits people rely on.