The Motive Model

Building a product takes more than one kind of expertise — someone to frame what’s worth making, someone to build it, someone to keep it coherent, someone to test it. A person is good at one. So no one advanced a product alone: you formed a team, and the work waited on whoever held the discipline you lacked. That waiting — not the typing — was the real cost of building.

AI changes who has to be present. A discipline can be codified — written down once — and run by an AI delegate that acts on your behalf, putting the engineer’s, designer’s, and tester’s judgment within your reach on demand. One person can now carry a product far past their own expertise without stopping to wait for the rest of the team.

With progress no longer gated by who’s in the room, a job title stops being a boundary. What defines your place on a team is no longer the job you hold but the direction you set right now — a motive (what you want from the work) pursued from your angle of expertise; the angle is what tells you which want is worth holding and how to judge it. You fulfill that direction by wielding a delegate you direct to act on your behalf. That shift — from title to motive — is the move everything else here is built on.

flowchart LR
  A["<b>Actor</b> — human<br/>motive · accountability"] -->|authors| S["<b>Delegation surface</b><br/>brief · contract · shape · corpus"]
  S -->|configures| D["<b>Delegate</b> — agent<br/>capacity · no motive of its own"]
  D -->|returns work| A
  A -.->|checks fidelity| D
  class A actor
  class S artifact
  class D delegate
  classDef actor stroke-width:2px
  classDef artifact stroke:#0d9488,stroke-width:1.75px
  classDef delegate stroke-dasharray:6 4

The whole model in one move: a human holds the motive and the accountability; the delegate is faithful capacity, with no motive of its own — a faithfulness the actor enforces by checking fidelity, not something assumed.

That is the whole model in one move. The rest of these pages unpack it.

The shift in one table

	Production scarce (pre-AI)	Production abundant (with AI)
Unit of a team	a fixed contribution — what a job title named	the direction — a motive, from your angle of expertise
A person	does one kind of thing well	spans several, via delegates
Producing the artifact	the scarce, defining act	cheap; no longer the bottleneck
The scarce thing	making the thing	deciding what to make, and judging it

What this model gives you

A vocabulary for building anything with AI, organized so the pieces don’t overlap. Read it as a path:

Positions Are Not Roles Start here. How your familiar job title cracks open into several motives.

The Four Actors The four motives every build needs: intend, generate, structure, accumulate.

Two Faces and the Gate Why everyone both produces and judges, and how ship/kill decisions actually work.

Delegates and Surfaces What a delegate is, how you hand it intent, and how you check it stayed faithful.

Variants and How People Grow Same motive, opposite training: where careers specialize.

Scenarios The model in motion, from a solo dev to an open-source team.

Recursion The same four motives build the product, the process, and the toolchain.

Glossary Every load-bearing term, defined once.

Why bother getting this right

Get the motives right and everything downstream follows — what each person needs to do, how the work plays out, what surfaces they hand to their delegates. Get them wrong — by treating an AI as a teammate with its own goals, or by organizing the team on a timeline instead of by motive — and the work overlaps, the hand-offs blur, and the confusion compounds.

The model has one load-bearing rule, stated up front so nothing later contradicts it: an AI is never an actor. A human holds the motive and the accountability; the AI is capacity the human wields. The next page shows where that line falls in work you already do.

Next: Positions are not roles →