Image to video is one of YouTube's first GenAI creation tools, using Gemini to simulate motion, transitions, and transformations of users' images directly within YouTube Shorts.
My Role
I led UX and visual design from vision to implementation. The project started with an idea I proposed: a dedicated animation mode within the YouTube creation flow. I built prototypes to pitch the concept to cross-functional partners and stakeholders, and it became one of the most critical projects on YouTube's creation roadmap.
Process
The team went through multiple rounds of iteration, constantly navigating the tension between short-term shipping constraints and long-term scalability. Every design decision had to work for today's milestone while leaving the door open for what comes next.
UX Strategy
Suggestions are generated by analyzing the user's image and spanning a spectrum from practical to playfu, ensuring the tool feels useful to someone making a quick edit and inspiring to someone who wants to experiment. The goal was to meet users wherever they are creatively, without requiring them to know what they want before they start.
Explorations
We designed multiple variations for how suggestions could surface, anticipating future features like open prompts, camera controls, and style transfer. The core design challenge was ambiguity, building a system flexible enough to sequence new capabilities naturally as they became available, without requiring a redesign each time.
