The Amplifier Model

The Amplifier is not a personality type you are born with. It is a posture you practise. This section lays out the three ideas the book is built on, in reference form, free to read.

In brief

The Amplifier Model has three parts. The three types describe how people default under a powerful new tool: the Dependent outsources thinking, the Avoider refuses to engage, the Amplifier thinks first and delegates deliberately. The eight-step framework is the smallest sequence of decisions that protects against both traps. The Amplifier Stack is what the same idea looks like fifteen years into a career.

The Amplifier is not someone who is good at AI. The Amplifier is someone who thinks clearly and uses AI inside that clarity.— The Amplifier, Appendix A: the core principle
The Three TypesDependent, Avoider, Amplifier — how each forms, what each costs, and why the two traps are not mirror images. The Eight-Step FrameworkThe full method: the eight steps, the Decision Filter, the Tool Selection Logic, two-step prompting and the iteration logic. The Amplifier StackFor anyone with a career built before AI: what to keep, what to add, what to double down on. GlossaryThe book's vocabulary, defined — from the Dependency Trap to the girl at the gate. FiguresThe book's diagrams, described and downloadable. Take the AMQLocate yourself in the typology in about twenty minutes.

Why a model, and not a list of tools

The word "AI" has grown so big it has stopped meaning anything. In the book, Vivek makes Arjun notice he has used it twenty times in fifteen minutes without being able to say what he means. The model exists to replace that formless word with a vocabulary: a way of seeing which trap a colleague is in, which trap you are in, and what the third path actually requires. Once the words exist, the room becomes visible. That is the first thing the model does, and it is why this section comes before the tools.

The second thing the model does is outlast the tools. ChatGPT, Claude, Gemini, Cursor, Midjourney, NotebookLM — every name in the book is a name for now. In two years half of them will have been replaced. The categories will not have. The framework is built on categories and on decisions a human makes, which is why it can be taught in 2026 and still hold in 2031.