The Cost of Intelligence
It falls to zero, and strategy doesn't follow it down. The price of running a model is collapsing; building an advantage on top of it is where the work moves.
Essays
The Judgment Line
Loop engineering for knowledge work, where you are the grader. What a loop actually is, why most of what gets sold as automation isn't one, and how to isolate the judgment only a human can supply.
The AI-Native Individual Contributor
When agents do the production, the individual contributor's work turns into routing, verification, and ownership — and that quietly breaks the way people used to learn the trade.
The Agentic Operating Model
When an agent is the default actor for incoming work and an orchestrator decides which one, the operating model is the whole game — and a wrong route in a regulated domain is a compliance event, not a bad answer.
How to Restructure for AI
Not org-chart tinkering. Structure is the design variable, most firms are pulling the wrong lever, and the evidence on flattening is older and more inconvenient than the pitch admits.
Borrowed Minds: A Model-Call Layer for AI Agents
A small CLI that lets one AI agent call other models and pull their answers back into the session — and the restraint that keeps it useful.
An Operating System for Judgment
What each part of one executive's AI operating system does, and why it helps — a walk through the map, box by box.