Skip to main content

Step 2 of 11

The AI Engagement Model

A framework for describing how a person works with AI, and for deciding how you want to work with it yourself.

Most advice about AI is advice about tools. Which model to use, which prompt to paste, which feature shipped this week. That advice ages badly, and it skips the part that actually varies: what the person does.

Two people can open the same model, ask about the same problem, and end the hour in completely different places. One has an answer. The other has an answer, a reason to believe it, and a better version of the question they started with. The difference is not the tool. It is the engagement.

The AI Engagement Model is a way of naming that difference precisely enough to teach it, measure it, and choose it.

How the pieces fit together

The model has four interacting layers and a fifth layer of outcomes, with a developmental loop that closes back on the first. Read from the bottom up, it says: the situation shapes what is possible, who you are shapes what you can sustain, what you believe shapes what you choose, and what you choose shapes what you get.

  1. Mechanism

    A person, a tool, a task and a context produce the interaction together. None of them determines it alone.

  2. Dispositions

    Durable tendencies that make wise engagement easier or harder. Five of them.

  3. Principles

    Rules for making sound engagement choices. Six of them.

  4. Behaviors

    The eight engagement modes you can actually observe in a conversation.

  5. Outcomes

    What the engagement strengthens or weakens: learning, judgment, agency, voice.

Becoming — the loop back, not a sixth layer

Across many engagements, repeated reflective behavior can strengthen the dispositions the chain started from. That return arrow is what the model calls Becoming. It is a loop over time, not a further layer, and no single conversation, mode count or score demonstrates that it has happened.

Quick check

Dispositions sit below behaviors in the diagram. What does that ordering mean?

The rest of this walkthrough takes about fifteen minutes. It covers why the framework exists, the theory that decides what the layers are and what prior work that theory rests on, what each layer contains, what the research has actually found so far, and how to use the model to govern your own AI use rather than just describe it.