While the guy in this AI video looks just like me, the clip was actually created using one of the new features available in the Google Gemini app: avatars. This digital recreation resembles the core features of OpenAI’s now-defunct Sora app. It is a digital version of you and can be inserted into AI videos. The avatars are powered by the company’s new Omni video model, and this feature is only available to subscribers.
I pay $20 a month for the Google AI Pro plan and quickly maxed out Gemini’s usage limits, which reset every 5 hours. I simply asked a few questions and created two 10-second clips featuring my avatar, before being asked to wait until later.
Video: Rhys Rogers
My first two glimpses of what the Omni could do to my likeness were when I was singing to Dino in San Francisco and surfing under the Golden Gate Bridge. I was impressed and horrified at the same time. The content was frustrating, with some jumbled moments and meaningless costumes, but that guy in the video was me. I used my fingers to zoom in on her face and really watched the mouth move. His teeth were slightly chipped, but otherwise he was Reese, right down to the chin fat.
Unlike OpenAI, which previously allowed users to decide whether they wanted others to create AI-powered videos using their photos, Google only allows adult users to create videos using their avatar.
It took about five minutes to set up my avatar through the Gemini app. The process involved sitting in a well-lit room, pointing my phone’s camera at my face, and reading out a series of two-digit numbers. Then I slowly looked to the right and turned my head to the left, and it was all over. Reece 2.0 was born and she’s ready to be my deepfake star. (Be mindful of what you’re wearing during this process, as what suits you will likely show up in AI generations, but more on that later.)
Let’s break down the Christmas clip frame by frame to unpack my feelings here. Full prompt: Make a video of me singing Happy Birthday to an old dinosaur on the top of the hill in Dolores Park.
AI generated clip by Rhys Rogers
The first second starts with Alfie pausing because even AI Reece has some ingrained habits. What initially catches the eye is the realistic setting. Instead of placing my avatar on an oversized hill in a random park, Google’s AI-powered video background looks more like the actual location. From the palm tree-lined sidewalks to Salesforce looming in the distance, it’s immediately clear which park was photographed here, even though the output isn’t perfect. It makes sense that a company known for mapping the planet could achieve this goal.
When the AI started singing, in a lower-pitched baritone than I could actually do, the first few bars sounded natural. I moved my hand up and down to the beat, like the conductor of a small band. Then, I stutter over the word “to,” and Jiminy cuts to a wider angle shot when the real chaos begins. A vanilla cake appears randomly, and I exhale a cloud of smoke to blow out the celebration candle. (Honestly, how rude of AI Reece to be. It’s not your special day.)
