Class Six
Today we expand the field of AI photography, then launch your final project. By the end of this class you will have a group, a lead character, and a production approach. The work begins now.
What We Are Doing
Expanding the Field: AI Photography Today
We revisit Shane Hulbert and then look at five more artists working with AI as a medium: Boris Eldagsen, Nouf Aljowaysir, Ben Millar Cole, Jake Elwes, and Philip Toledano. Six different positions on what AI can do in photography and why it matters.
Project Brief: The Monologue and the World
The full brief for your final project: a monologue video, five still images, and a production book. What is required, how groups work, the two production options, language choices, and what the final presentation looks like.
Group Formation and Character Selection
Form your group of up to four. Share your characters with each other. Choose which character leads the project and begin mapping how the other characters exist within that world.
Production Planning Begins
The production plan does not start at home. It starts here. Your group should leave today with a concept for the monologue video, a rough plan for at least two images, and a shared folder started for the production book.
Expanding the Field
In Class 4 we looked at Shane Hulbert as an example of AI used as a primary image-making medium: characters constructed entirely in AI, with the same care and depth as any conventionally photographed tableau. That was one position. There are others. Below are five more artists working with AI in photography, each with a distinct argument about what the medium can do. Know where you stand in relation to them before you choose your approach for the project.
Shane Hulbert
The baseline position we have already encountered. Hulbert replaces documentary objectivity with synthetic construction. AI as the primary medium, not a filter. The images look like photographs. They are not photographs.
shanehulbert.com →Boris Eldagsen
Eldagsen submitted an AI-generated image to the Sony World Photography Award 2023 and won. Then he rejected the prize. His argument: AI images should not compete in photography categories because they are a different medium. The provocation was the point. If you cannot tell the difference, does the difference matter?
eldagsen.com →Nouf Aljowaysir
Aljowaysir's work interrogates what is inside an AI's training data: specifically, how colonial-era photographic archives shaped what AI “knows” about non-Western identities. When you use AI to generate your character, whose idea of a person is the model drawing on?
noufaljowaysir.com →Ben Millar Cole
Cole creates AI-generated images that look like photographs from a collection that does not exist. They have the grain, the fading, the compositional language of real archival material. No person in them ever existed. No photographer ever pressed a shutter. Photography as archaeology of a memory that never happened.
benmillarcole.com →Jake Elwes
Elwes trains AI systems on datasets he builds himself, deliberately and with a specific politics. His argument is direct: AI reflects what you put into it. If you build a dataset with intention and specificity, the AI produces something different. The choices you make about what to feed in are not neutral.
jakeelwes.com →Philip Toledano
Toledano uses AI to generate photographs of historical moments that never happened, and to reconstruct Robert Capa's lost D-Day photographs, destroyed in processing. His argument: photography has always been a selective and constructed version of truth. AI makes that construction visible rather than hidden.
America → England → America + →Six Positions
Taken together, these artists give you a map of where AI sits in contemporary photography practice. You do not need to align with any one position. But you should be able to say, clearly, where your own approach sits in relation to them.
The Project Brief
This is the brief for your final project. Read it carefully. Everything your group produces from now until the final presentation class is part of this project.
The Monologue and the World
Your group will produce three things: a monologue video, five still images, and a production book. They form one coherent body of work, not three separate tasks. The video is the interior voice. The images are the world that voice inhabits. The book is the thinking that made both possible.
- The monologue video Your group’s lead character speaks their interior monologue. The same monologue developed in Class 5 — filmed or generated in AI. Length: approximately 1 to 2 minutes. Shown on screen at the final presentation.
- Five still images A body of work connected to the monologue. The world the character inhabits. Other characters in their story. Moments implied by or connected to the monologue text. A3 or larger, photographic quality prints, displayed in the room at the final presentation.
- The production book Everything that went into making the work: 30 Questions for each character, monologue scripts, mood boards, drawings, storyboards, call sheets, production plans, reference images, test images, rejected ideas, revised drafts. A4, bound, presented alongside the prints.
Groups and Characters
Groups of up to four. Each student has already built their own character through the 30 Questions and monologue work in Classes 4 and 5. The group chooses one character as the lead for the monologue video. The other group members’ characters are not discarded — they are integrated into the five still images as secondary figures in the story. Their relationships, tensions, and connections shape what the images are about.
The lead character’s world is the primary world of the project. The other characters exist within it. Whose world has the most to say?
Two Production Options
The monologue video and the five images can be produced conventionally, using AI, or as a combination of both. There is no hierarchy between the two approaches. Both require the same depth of character, the same production planning, and the same specificity of intention.
Language
The monologue can be delivered in any language: English, Chinese, or any other. The language choice is itself a character decision. What language does this person think in? What does the language carry that another language does not?
The Final Presentation
In the final class your group presents all three deliverables together. The video plays on screen. The prints are displayed in the room. The production book is presented alongside them. The work is assessed as a complete body of work. Expect to have approximately one week of production time, with work in progress reviewed in the next class.
Forming Your Group
Groups of up to four. Once formed, share your characters with each other: name, age, world, occupation, one specific detail, and the opening lines of your monologue. Listen for connections. Whose character has the most visual potential? Whose world is most developed? Whose monologue is most complete?
Choose one character as the lead. This is the character who speaks the monologue in the video and around whom the five images are built. The other characters become part of that world. Then ask: how do the other characters connect to this person? Are they colleagues, family, strangers, intimates? Do they know each other? What is the nature of their relationship?
Groups do not need to be friends. They need to have characters that could plausibly share a world. Tension between characters is often more interesting than similarity.
Production Planning Begins
This is not a homework task. It starts now. Your group should leave this class with clear answers to the questions below. A group with a strong concept and a shared folder is in a much stronger position than a group that spends the first three days of production week still deciding what to make.
Where You Should Be
The next class will review work in progress. You should arrive with material to show, not just ideas to describe. Production time is limited. The groups who are furthest along at the work in progress review are the groups who have been working since today.
- Group formed, up to four students
- Lead character chosen: the character who will speak the monologue in the video
- A clear sense of how the other group members’ characters connect to the lead character’s world
- Production approach decided: conventional, AI, or combination
- Language of the monologue decided: English, Chinese, or other
- A starting concept for the monologue video (location or visual style, approximate length)
- A rough plan for at least two of the five still images
- A shared folder or document started for the production book, with the 30 Questions and monologue drafts already in it
- At least one test image, sketch, or location reference produced before the next class