Google I/O: Gemini Omni AI Debuts with Advanced Video Tools

Google has unveiled a new AI world model called Gemini Omni at Google I/O. This model features advanced AI video generation capabilities. The first release in this family is Gemini Omni Flash. What's special about Omni is its multi-modal input and output. Users can input text, audio, images, and video. Omni then generates unique, interactive worlds using its real-world knowledge. Google says Omni will produce videos with more accurate physics and realistic content. It can also understand contextual prompts, like historical facts, for greater accuracy. Users can also edit videos conversationally with Omni, changing backgrounds, styles, angles, or specific details. Omni will also allow users to create digital avatars, a feature still under testing for responsible launch. All generated videos will have a SynthID watermark for verification. Gemini Omni Flash is rolling out today to paid Google AI subscribers and will launch later this week in YouTube Shorts and the YouTube Create app at no cost. This development aims to bring more sophisticated and accessible AI video tools to a wider audience.