Gemini Omni: Clone yourself with AI in under 15 minutes

Key takeaways

AI video tools like Google Flow and Gemini Omni can significantly reduce the time and effort needed to produce personalized video content.
Creating digital avatars of oneself is accessible, initial results are promising, but current technology still has limitations with emotions and realistic physics.
These tools are opening new creative possibilities, enabling people without video production skills to generate professional-looking content quickly.
Rapid experimentation with AI can lead to unexpected, humorous, and sometimes uncanny results, highlighting the ongoing improvements in AI realism.
Integration of AI-generated content into workflows can streamline content creation, making it easier to produce storytelling videos, marketing, or personal projects.

Summary

Introduction to AI Avatar Creation and Experimentation

The episode begins with the host attempting to create a video avatar of themselves using Google Flow and Gemini Omni's new video generation model. The process involves scanning their face with a phone, generating an avatar, and building a storyboard for a podcast hype video. The host expresses curiosity and some uncertainty about the process's success, highlighting the novelty and experimental nature of the project.

Using Google Flow and Gemini Omni for Video Generation

The host explores Google Flow's avatar creation feature by scanning their face and creating a digital persona. They discuss the potential of multimodal AI tools to unlock creative possibilities, emphasizing that these tools can enable individuals with no video production skills to generate high-quality content. The process involves brainstorming, storyboarding, and attempting to generate scenes with the avatar, often facing challenges like referencing accuracy and technical hiccups.

Challenges and Improvements in AI Video Generation

Throughout the experiment, the host encounters issues such as generating images instead of videos and inconsistencies in avatar realism. They showcase the AI's uncanny valley moments, where facial expressions and emotions appear unnatural. Despite these hiccups, they highlight the potential and rapid progress of current AI video tools, noting that even imperfect results can be achieved with minimal effort and time.

Creating and Stitching the AI-Generated Video Content

The host stitches together multiple scene clips into a full hype video using browser-based editing tools. This process is quick and accessible, demonstrating how AI can streamline video production. The final product includes scenes of the avatar speaking and performing actions, with some humorous and uncanny moments. The result showcases the rapid creation of a cohesive video from AI-generated scenes breaking the traditional barriers of video editing.

Reflections and Future Possibilities of AI Avatars

The host reflects on the experience, noting that the generated avatar more than 50% resembles their face and can produce convincing results with further refinement. They acknowledge imperfections like uncanny expressions but also recognize the significant time savings and creative potential. The episode closes with enthusiasm for future experimentation and encouragement for others to explore AI avatar creation, emphasizing that even minimal effort can yield impressive results.

Show notes

In this experimental episode, I document my real-time attempt to create an AI avatar of myself using Google Flow and the new Gemini Omni video generation model. I walk through the entire process—from scanning my face with my phone to generating a complete one-minute hype video for the podcast, all in about 15 minutes.

What you’ll learn:

How to create an AI avatar using Google Flow in under five minutes
Why video AI tools unlock creative possibilities for people with zero video production skills
The step-by-step process of generating a full storyboard using AI as your creative producer
How to use character consistency features to generate multiple video scenes with the same avatar
The uncanny-valley moments you’ll encounter when your AI clone doesn’t quite nail emotions or physics
How to stitch together AI-generated scenes into a complete video using built-in editing tools

—

Brought to you by:

Merge—Connective infrastructure for production AI

Jira Product Discovery—Prioritize with insights, build with confidence

—

In this episode, we cover:

(00:00) Getting started with Google Flow and Gemini Omni

(01:38) The avatar creation process: scanning and photo capture

(02:55) Using Flow to brainstorm a hype video storyboard

(06:59) Generating the first video scene with the avatar

(08:41) Troubleshooting: accidentally generating images instead of videos

(09:32) Generating all seven scenes for the complete video