YouTube Unveils Upgraded AI for Shorts
YouTube has unveiled new generative AI tools for Shorts creators during its annual “Made on YouTube” live event, reports TechCrunch. The tools include a custom version of Google’s text-to-video model, Video (Vi) 3, a new remixing tool, and an Edit with AI feature.
The custom version, named Vi 3 Fast, can generate videos at 480p resolution with low latency. “For the first time, it allows creators to produce video clips with sound,” YouTube said. The feature is initially rolling out in the United States, United Kingdom, Canada, Australia, and New Zealand, with plans to expand to other regions later.
Vi 3 also introduces new capabilities for Shorts. Creators can animate still images using motions learned from videos—for example, making a person in a photo replicate dance moves from a video. Additionally, videos can now incorporate pop art or origami-style effects, and new characters or props can be added simply by providing text descriptions.
The new remixing tool allows creators to convert approved video dialogues into engaging soundtracks for use in other Shorts.
YouTube stated that these features will be gradually rolled out over the coming months.
