Method / not a formula
Enter the idea
The process uses generative video to move from an abstract proposition to an inhabited situation—and then back to a sharper real-world question.
- 01Name the propositionReduce the idea to the change it hopes to create.
- 02Assign a rolePlace the cameo as user, guide, maker, witness, or critic.
- 03Build the situationAdd place, props, other people, sequence, and consequence.
- 04Watch for frictionNotice what feels persuasive, missing, comic, false, or unexpectedly alive.
- 05Return to realityConvert the reaction into a design question or next prototype.
The prompt is a casting decision
A useful prompt does more than describe a beautiful shot. It gives the cameo a reason to be there.
“A person in a garden” is an image. “A neighborhood arborist leads families on a tree walk, pausing beside a numbered stop while someone opens the guide” is a situation. The second prompt contains relationships, tasks, implied interfaces, and a standard by which the scene can feel coherent or wrong.
The role also changes the questions. As founder, the cameo tests whether the proposition can be explained. As participant, it tests pacing and motivation. As maker, it reveals the physical production sequence. As customer, it makes delight and hesitation visible. As comic performer, it exposes assumptions through exaggeration.
Keep one variable familiar
The recurring face is a control point inside a highly variable medium.
Generative video changes everything at once: camera, setting, action, wardrobe, material, other characters, and sometimes basic physics. Reusing @jrb20 makes at least one element legible across that instability. It supports comparison between worlds and creates continuity across otherwise unrelated projects.
What position does the cameo occupy in relation to the idea?
What sequence of gestures turns an object into an experience?
Who else must be present for the scene to make sense?
What changes after the imagined product or service is used?
This is not user research
A synthetic scene cannot tell us what other people need, whether they can use a product, or how a real community will respond.
Generated people do not consent, improvise from lived experience, resist a premise, or reveal needs the prompt author cannot imagine. The cameo is especially limited because it centers the maker's own body and expectations. It risks turning a project into a flattering launch film.
The responsible use is narrower: treat the clip as a speculative sketch. Let it clarify what to build, what to ask, whom to involve, and which assumptions need real-world contact. It is preparation for research—not a replacement for it.
From reaction to next move
The most productive question after a clip is not “Is this good?” It is “What did seeing it make newly discussable?”
A tree-walk video may reveal that the guide needs a social role, not just a map. A music-card scene may show that the pleasure lies in choosing and handling, not merely scanning. A VR archive may make density feel impressive but exhausting. A self-driving scene may transform abstract trust into a visible bodily reaction.
The clip has done its job when it changes the next action: make a paper prototype, revise the sequence, invite a subject-matter expert, test a physical scale, remove a feature, or stage the situation with real people.