Notes · Field log
Lessons from the field.
Short, practical notes from real builds, what broke, what fixed it, and what I'd tell you before you hit the same wall. Updated as the work teaches me.
01
Directing AI, not just prompting it
Everyone can get an answer out of an AI now, so the answer was never the value. My rules for when to lean on it, when a human stays in the loop, and how I keep accountability for every number that reaches a decision.
Read →
02
Thirteen iterations: what a 53-million-candidate bug taught me
The most expensive defect in my Center of Gravity tool was invisible in code review, it only appeared when the tool met real data. Notes on iterative execution, primary sources, and directing AI as an engineering partner.
Read →
More notes in progress, this log grows as the work does.
Working through something similar?
If one of these notes maps onto a problem you're wrestling with, I'd genuinely like to hear about it.