On what horizon to judge a system
A year spent from one client to the next. You see up close how an AI budget line is defended or collapses in committee, and on what horizon a system should be judged.
A whole year spent from one client to the next, between a group digital factory and a pharmaceutical accelerator. Changing client every six months teaches one thing: the real question about a system is what it becomes once you have left. The year ahead should bring longer engagements, and I notice that is what I am after.
It treats AI as an investment to justify, with abandonment criteria written in advance. The most useful chapter explains how to stop a project without dressing it up as a success.
It proposes judging a system by what it becomes once you stop tending it. That is exactly the test of a handover, and I have found none better since.
A century of perspective on complexity, written short. What a practitioner keeps: the urge to simplify a system is nearly always the urge to ignore part of it.