YouTube Shorts Integrates Veo 2 for AI Video Background and Clip Generation

At last year's Made on YouTube event, YouTube released a high-profile update to its Dream Screen feature. The feature allows users to create unique AI backgrounds for their Shorts with just a simple text prompt. From fun ways to learn to creative remixes, YouTube emphasized that many users are already using Dream Screen to bring their creative visions to life.

Now, Dream Screen is getting a major upgrade that integrates Google DeepMind's latest video generation models Veo 2YouTube also announced a new feature powered by Veo 2: the ability to generate standalone video clips. Users can seamlessly add these clips to any Shorts video.

YouTube points out that when users are struggling with a lack of scene-specific material, or are eager to bring their imagination to life to tell a unique story, it's easier and more convenient to simply enter text prompts to generate a video clip that fits the narrative perfectly, or create a world of out-of-this-world content.

YouTube says that Veo 2 works in tandem with Dream Screen to generate top-quality videos that cover a wide range of topics and styles to better match a user's vision. In addition, YouTube claims that Dream Screen has been optimized for faster video generation, and that Veo 2 understands the laws of real-world physics and human movement more accurately, resulting in a more detailed and lifelike output. Users can even specify video styles, shots, or cinematic effects, making Dream Screen an easy and fun way to express yourself.

• To create a video background for Shorts, users can select the Green Screen and Dream Screen features in the Shorts imager to generate a video background with text prompts and selected images.
• To generate and add a video clip to Shorts, users can click Add in the Shorts imager to open the media selector, select Create and enter a text prompt, select an image and video length, and then create a video clip.

To highlight their responsibility in AI innovation, YouTube emphasized that they used SynthID watermarks and clear labels to mark this AI-generated content.

These features are currently available to all users in the U.S., Canada, Australia and New Zealand, with plans to expand to additional countries in the future.

YouTube is looking forward to bringing the upgraded features to users, helping them bring their creative visions to life at their fingertips and ushering in a new era of video creation.