Revival
Acting as creative director, I developed the narrative and storyboard while working closely with generative AI through iterative prompting to translate imagination into imagery. The project reframes AI not as a replacement for human creativity, but as a tool that expands it, allowing humans to imagine and build more regenerative futures.
The film centers on the theme of collaboration between humans and AI, expressed through generative AI–driven storytelling. While the narrative and storyboard were entirely my own, AI became a collaborative tool throughout the visual development process. I generated images through iterative AI prompting, refining, and rewriting prompts until the visuals aligned with my intent. This process involved numerous iterations, as the AI rarely produced exactly what I envisioned on the first attempt.
"How might human vision and artificial intelligence unite to regenerate nature and rebuild a thriving planet?"
Skills
Speculative Future Worldbuilding, Environmental & Climate Storytelling, Video Editing, Film Producing, Screenwriting, Concept Development & Visual Storyboarding, Directing
Set within the “preferred future” of the Cones of Possibility framework, this short film envisions a possible solution to this question.
Identify Signals & Trends
Mapping Future Scenarios
Developping Narratives
Initially, I envisioned an A.I. database designed to archive and locate seed banks after a global disaster eliminated all plant life on Earth. Humanity would depend on this system to restore the planet using preserved seeds.
After receiving critique, I recognized that the idea was too complex and that audiences might struggle to grasp the database’s role. I was advised to focus more on storytelling and worldbuilding, so I returned to developing the narrative—refining the setting and characters to create a clearer, more engaging story. Once the story felt solid, I began storyboarding the project.
I experimented with multiple generative AI tools, including Reve, Invoke, Midjourney, and Flick.art. Many of these tools produced results that felt unrealistic or overly artificial, making it challenging to create visuals that did not immediately read as AI-generated. Some of the Image generation failures are included on the left.
Each image required numerous iterations. Translating the nuances of my future world into AI-generated imagery was difficult, and achieving the level of realism I envisioned took extensive refinement. The process became a highly collaborative exchange between myself and AI. However, all editing, scriptwriting, and creative direction were completed by me.