Fear - an AI short film.

THE CHALLENGE
The fastest way to understand a new generation of tools is to ship something real with them. “Fear” was a self-initiated brief with two jobs: make a film that stands on its own as a piece of my self-expression work, and use it to pressure-test today's AI image and video models end to end, not in demos, but in a finished, controlled, on-brand result.

THE APPROACH
The film is a painterly motion graphics piece with a deliberately locked visual language: oil and drypoint etching, a near-black palette, and a single crimson red accent. The constraint was the point. Holding one exact look across an entirely AI-generated pipeline is where these tools usually fall apart, so that became the test. Every frame was iterated to stay inside the style rather than letting the model wander. The film opens and closes on the eyes, a bookend that gives the piece its shape.

I used to think of fear as something to beat. Now I think it's just an old part of me that never got the memo the danger passed. This film is about that, not fighting it, just finally turning around and seeing it. Most of our suffering doesn't come from what happens to us, it comes from resisting what already has. Acceptance doesn't make the fear disappear. It just means you stop running from something that was only ever you.

THE PIPELINE
The piece was produced across current AI tools as a connected workflow. Generation, motion, and refinement passed between Nano Banana Pro, GPT Image 2, Seedance 2.0, Kling 3.0, and VEO 3.1, with the studio's hand on style and continuity throughout. Voiceover ran through ElevenLabs and Seed Audio 1.0. Final edit, SFX, and music were handled by hand to pull it all into one cohesive piece rather than a stitched-together demo reel. The takeaway is the real deliverable: a clear, hard-won read on where these tools excel, where they break, and how to direct them toward a fixed creative target.