2023
Role: Creative Technologist
Tools: Stable Diffusion, ChatGPT, ElevanLabs, Runway,
Pika Labs, AI sound generation tools
<Wait for an Update> is an experimental AI-generated movie trailer created in 2023. The project explored how realistic and cinematic a film could feel using almost entirely generative AI, while testing the possibilities and limitations of the technology at the time.
CONCEPT
Wait for an Update imagines a future where AI-generated people are so realistic that it becomes difficult to distinguish them from real humans. The project explores what happens when we begin to depend on AI that can imitate human behavior but is still limited by its programming. In this imagined world, AI characters can communicate like humans until they reach the limits of their knowledge, without an update, they become stuck, repeat themselves, and are unable to continue.
CREATIVE PROCESS
I started by developing the story and breaking the script into individual scenes. I translated each scene into detailed text descriptions and generated the characters, objects, and environments entirely from prompts, without using reference images.
Once I had the still images, I moved into animation. Keyframes became an important part of the process: I selected images as visual start and end points and experimented with up to five transitions between them to create movement and continuity.
I also focused heavily on timing and pacing, adjusting the duration of scenes and transitions to build tension and make independently generated visuals feel like one continuous trailer. Finally, I combined the animation, dialogue, sound, and music into the final sequence.
EXPERIMENT &FINDING
One of the most surprising results was how closely AI could interpret visual descriptions. Even without reference images, it could generate people, objects, and environments that were often close to what I had imagined.
At the same time, the limitations were very visible. In 2023, AI struggled significantly with text inside images, often creating shapes that resembled letters rather than readable words. Visual consistency and transitions also required experimentation and careful selection.
The process showed that while AI could generate the individual pieces, creative direction, selection, keyframing, timing, and storytelling were essential to turn them into a coherent cinematic experience.