A guide for directors who make commercials with AI, from the concept to delivery, with every tool in its own complete section. It opens with a letter to my fellow directors.

This book takes a commercial from the first idea to the delivered cut, in the order the work happens, and teaches each AI tool once, in full, in its own section. This page tells you how the book is laid out, how to read an example, and which four decisions to settle before the fi

One picture and one table. The picture shows the stages of an AI commercial, the four gates that stop the work and the three records that keep it honest. The table says which tool does which job and which chapter teaches it.

You usually know the job before you know the chapter. This finder lists 162 jobs a director actually has on an AI commercial, in the order a production meets them, and sends each one to the section that teaches it, with the section's title beside its number.

On an AI commercial there is no camera to place, no set to light and no take you can watch being made. There is a model that reads what you write, looks at what you attach, and hands back pictures and sounds that nobody shot. This chapter describes what that changes: what the mac

A prompt is a brief for a reader who has never met you, reads it once, takes every word literally and forgets it the moment the job is done. This chapter teaches the principles that hold for every model: which work belongs to the words and which to the inputs you attach, the two

A generator makes the film, and what makes the film worth its credits is decided earlier, on paper: what it is about, what it refuses to do, and the one device that turns a truth about people into something an audience will repeat. This chapter teaches how to get from a brief to

A treatment has three readers now. The client must feel the film and buy it; the crew must know what to build, light and cast; and a model must eventually be told, literally, what is in each frame. This chapter teaches one document that serves all three: how to turn every feeling

A shot is a decision: what the audience must see, from where, for how long, and why this shot and not another. On a set, a good crew fills the gaps you leave. A model fills them too, with whatever it has seen most often. This chapter teaches the craft that leaves no gaps: the job

Light tells the audience where they are, what time it is, what to look at and how to feel, before anyone speaks. A model lights every frame whether you ask or not, and unasked it lights the way every generated image is lit: soft, even and from nowhere. This chapter teaches light

A generated room looks built because nobody has ever lived in it. Its surfaces have never been touched, its objects belong to no one, its clothes were never sewn, and its products float in their own light. This chapter teaches the production designer's answer: wear where hands go

On a set, planning protects the shooting day. An AI production has no shooting day: it has hundreds of small jobs, each cheap, each starting from nothing, each made by a model that remembers none of the others. Planning is what turns them into one film. This chapter teaches the d

Every film here starts as a still. Before anything moves, speaks or is cut, someone has to make one frame that is right: the right face, the right room, the right instant, lit the way the film is lit. This chapter teaches what a finished still is, whichever tool makes it: the twe

A finished still is seldom made from words alone. It rests on a face that must stay the same face, a room that must stay the same room, a product that must stay the client's product, a sign that must say exactly what was approved, and a signed frame that must change in one respec

Nano Banana Pro is Google's image model and the main stills camera of this book. It makes the hero frames of a commercial, the first picture of a character, the keyframes that the video models animate, and the repairs that save a frame that is nearly right. It reads a prompt once

GPT Image 2.5 is the second stills camera. Nano Banana Pro carries a signed face and the photographic keyframe; this camera makes what the first does not: an element on a transparent background, a frame led by text or layout, a packaging or key-visual study, and a bounded edit on

A still is one decision. A moving shot is two: what must survive from the pictures and words you already own, and what the model is allowed to do with them. This chapter teaches the choice every shot in Part 3 begins with: the operations, the one-line record of the choice, when t

A moving shot is a signed picture, a paragraph of words and a set of settings, each owning its own part of the job. This chapter teaches the paragraph and what surrounds it: what the frame already carries, the seven things a finished prompt says and in what order, how to size a c

A generated actor has no inner life to draw on. It draws what you describe, and it turns feelings into faces pulling expressions. This chapter teaches what makes a person believable on screen and how to get it from a model: one visible event for each beat, a task and a few cues,

Kling is the motion camera you reach for when the move is the shot: a dolly, a track, a crane, a push on a face or a label, or one small action from a frame you have already signed. This chapter teaches Kling 3.0 as it runs today through the Higgsfield connector: what it reads, h