Research notes.
What we're reading in the AI-integration field, what we make of it, and how it changes how we run engagements. Selected, not periodic.
Each note cites the primary source. The full bibliography is at Sources.
Reaction · May 19, 2026
Anthropic's 4D AI Fluency model — what it gets right, and where it stops
The 4D model captures the cognitive shift behind effective AI use better than most. The piece it doesn't address is what happens when a fluent individual sits inside an organization that hasn't done the adoption work.
Read the noteSynthesis · May 19, 2026
What Brynjolfsson 2025 means for who you train first
The empirics on who actually gains from AI assistance are sharper than the conventional pilot heuristic admits, and they argue for sequencing training the opposite of how most rollouts approach it.
Read the notePosition · May 19, 2026
Why we don't pilot with the most senior team
The conventional advice is to start AI pilots with the most senior cohort to build executive buy-in. The data says that cohort produces the least productivity gain, the most quality risk, and the highest visibility cost. We do the opposite.
Read the noteWant the next note in your inbox?
We publish selectively, not periodically. The simplest way to follow the work is the discovery call — it lets us tailor what we send.