Two things decide how a team behaves: what it believes to be true, and what it has agreed to do.

In 1740 the Scottish philosopher David Hume noticed that writers on morality would set out at length how the world is, and then start telling people how they ought to behave as though the second followed from the first. It does not. No description of a situation, however complete, has ever told anybody what to do about it. Most teams carry both of those things and have written down neither, and most companies are now handing real work to software with exactly the same gap in it.

Constitutional Leadership

A programme I am building for executive teams and boards, on those two things and the distance between them. It works on the map first, an honest account of where the team actually stands, and then on the rule, the written commitment about how it will act, because a team that has agreed only one of the two will keep being surprised by its own decisions.

AI Governance

The same gap, arriving faster and with nobody in the room. Businesses are handing real work to AI agents that book, buy, approve, message and decide, and most of those agents have been given a detailed picture of how the business runs without any written rule about what they may do inside it. I write about what governing an estate of them takes: an explicit, testable statement of what an agent may do, who decides, how it is watched, and what it may never do whatever the circumstance. The agents are arriving faster than the rules, and considerably faster than the regulation.

Both strands publish here and on LinkedIn. If either speaks to a room you sit in, the door opens soon, and the letters will tell you when.