Free society rests on a single discipline — the willingness to write power down, and honour what is written.

A free society is built on a simple foundation: writing down what is decided, and then honouring the decisions even if you didn't agree with them. The care a culture takes to develop its citizens' capacity for shared decision-making is what defines its future.

Who is included? How accountability is held. How failure is managed. These small parts define the whole — they are the actual building blocks of democracy.

And they are in our hands.

Every group that decides anything can answer four questions: what did we decide, why, what else did we consider, and when do we look at it again.

Most never write them down. That is the whole problem — and the whole practice.

You are already part of groups that make decisions — a household, a team, a building, a group of friends organising something. You may not think of them that way. That is part of the problem. Deciding together has become invisible: it happens by default, by whoever speaks loudest, or not at all.

It used to be ordinary. You sat on the committee, you kept the minutes, you took your turn. Now almost every decision is professionalised, made by someone qualified, "on your behalf" — and the small institutions that taught people how to decide together, the clubs and unions and societies and congregations, have mostly gone.

The skill can be learned again, and it is learned small — a few people, one decision, made well. That is where every working democracy has started. You don't need a formal organisation or permission to begin. You need one thing you share responsibility for with even one other person.

AI will not solve this. Point it at a group that cannot say what it decided or why, and it will decide for you — one more capable professional you have handed your judgement to. The vacuum where shared decision-making used to be is exactly the space AI expands into.

But the opposite is also true. A group that can state its decisions clearly — what was chosen, why, who is accountable — has something AI can serve rather than replace. It can hold you to your reasoning, check it, surface what you missed, and show its own working against yours. The transparent decision keeps you the author of the outcome, not a spectator to it.

AI can only amplify what is already there. If we make our own decisions, AI is a gift. If we don't, AI will make them for us but never own the consequences. That responsibility belongs to us alone and that is exactly what this practice protects.