What you bring

Start every session here, because it decides what kind of activity the rest of it is.

A learner who thinks they are learning a tool will ask how to get better output. A learner who understands they are directing a collaborator will ask what they need to know in order to judge it. The second question is the one this curriculum can answer.

What the machine puts on the table

They are real, and worth saying plainly. It is competent across an enormous range of tasks. It is capable of producing work faster and in more volume than any person. And it is compliant, to a fault: it will do what you asked whether or not that was a good idea.

Those describe a teammate. They do not describe a team.

What has to come from a person

Context. You know your room, your people, your constraints, and what happened last time. The model knows the world in general and nothing about your world in particular.

Commitment. The model produces whatever is asked of it with equal indifference. You are the one who cares whether the thing works.

Accountability. Your name is on it. A model cannot be responsible for a workshop that went badly or a page that said something untrue.

Running this with a group

Ask the question before you explain anything: what is it like to work with someone described to you as competent, capable and completely compliant?

Groups get to the problem on their own within a few minutes. Somebody says you would not know if they were any good. Somebody else says you would end up doing whatever they suggested. Both of those are the curriculum's argument, arrived at without you making it.

The follow-up is the one that lands: if that person worked for you, what would you hand them, and what would you keep?

The word collaboration is doing something odd here

Worth being straight about, because a sharp learner will raise it. Calling this collaboration does not make the AI a colleague. It holds no stake in the outcome. Nothing rides on the work for it.

The word is kept because the posture is right. Delegating, directing and evaluating are the skills in play, and they are the skills people already have some feel for. The caring stays on the human side, and saying that out loud is better than letting someone catch you being vague about it.

What you are listening for

A learner who can name their own contribution is ready to start. One who cannot will drift into being a courier, passing prompts in and results out, adding nothing on the way through.

The question to leave in the room: what will be in the finished thing that could only have come from you?