Writing
I write about building products, leading teams, and the decisions that shape both — especially now that AI is changing what all of those words mean. My writing is shaped by my experience across design, product, and leadership — twenty years of real work, and where that journey has taken me. Some of these pieces started as internal memos or working notes. They've been edited for publication, but the thinking came from the work.
Start here
Two gates decide most of your build quality
Quality isn't an implementation problem. It's a specification problem and a verification problem.
April 28, 2026 · 5 min read
Confidence scales. Evidence doesn't.
Intelligence improves the quality of the argument without improving the quality of the proof.
March 28, 2026 · 6 min read
0→1 work requires a different mindset
The phrase "zero to one" has been butchered. What matters is the proof contract, not the timeline.
March 24, 2026 · 4 min read
All articles
Two gates decide most of your build quality
Quality isn't an implementation problem. It's a specification problem and a verification problem.
April 28, 2026 · 5 min read
Responsibilities × Activities × Outputs
Titles sound precise until you try to use them. A clearer way to define a role is to separate what someone owns, does, and produces.
April 27, 2026 · 5 min read
Most of what's called "building with AI" is prototyping
"Build X in 10 prompts" is a sales pitch, not a methodology.
April 22, 2026 · 4 min read
If it only lives in chat, it does not exist
A decision that only lives in chat is a decision that hasn't been made yet.
April 17, 2026 · 5 min read
AI drift is real, and hallucinations come with it
AI systems don't fail only by making things up. They also fail by staying plausible after the truth has changed.
April 13, 2026 · 5 min read
A feature is not a product
A capability placed inside a buyer context is a product. A capability without one is a demo.
April 8, 2026 · 4 min read
Confidence scales. Evidence doesn't.
Intelligence improves the quality of the argument without improving the quality of the proof.
March 28, 2026 · 6 min read
0→1 work requires a different mindset
The phrase "zero to one" has been butchered. What matters is the proof contract, not the timeline.
March 24, 2026 · 4 min read