Journal
Notes on picking what to build.
Evidence, experiments and the occasional strong opinion about founder–market fit.
Manifesto · Latest
You can build anything. That’s the problem.
AI made building cheap. It didn’t make choosing easier — it made it the whole game.
September 25, 2026 · 4 min readEvidence
Receipts, not titles: what counts as evidence of fit
A practical guide to the proof that actually predicts whether you’ll win a market.
September 24, 2026 · 5 min readPlaybookThe two-week probe: test an idea before it eats a year
A Fit Report ends with an experiment, not a verdict. Here’s how to run it.
September 23, 2026 · 4 min read