Reference pieces and working ideas — written to be useful, not theoretical.
- 0110 May 2026Transformation
Aim Lower, Ship Sooner
The biggest aggregate value isn't in the big systems being replaced — it's in the thousands of small processes living outside them. AI now makes the build trivially fast. Aim lower, ship sooner, compound.
2 min read - 023 May 2026Strategy
Finding the Non-Obvious AI Use Case
AI opportunities — classical ML or GenAI — don't announce themselves. Three places worth looking: the human bottleneck, the always-skipped work, and the conservative gap. The obvious ones are already crowded.
2 min read - 0312 April 2026Architecture
AI Where Humans Slow Down
A working deterministic system doesn't want a model inside it. But every system has human layers — review, classification, drafting, judgement — and that's where AI belongs.
2 min read - 0415 March 2026Methodology
The Imagination Problem
The standard objection to GenAI projects is that the data foundation isn't there. That conclusion gets the constraint wrong — the real barrier is leadership willing to imagine the use case.
2 min read