The highlight of YouTube’s Made On YouTube event on Wednesday morning was undoubtedly artificial intelligence. The platform unveiled its integration of Google DeepMind’s AI video generation model, Veo, into YouTube Shorts, empowering creators to produce high-quality backgrounds and six-second clips effortlessly.
Unveiled at Google’s I/O 2024 developer conference, Veo is an advanced video generation model designed to compete with OpenAI’s Sora, as well as other notable video generation tools like Pika, Runway, and Irreverent Labs. This innovative technology can generate 1080p video clips in a variety of cinematic styles.
Veo is set to enhance YouTube’s existing AI feature, “Dream Screen,” which debuted in 2023. Dream Screen allowed creators to generate backgrounds in Shorts using text prompts, but with Veo, the video background creation process will be significantly improved, resulting in more impressive clips. A standout feature of Veo is its ability to remix and edit previously generated footage.
For the first time, creators will be able to produce six-second standalone video clips for Shorts. When users click “Create” and input a prompt, Dream Screen will generate four images, enabling them to choose one to transform into a video. This new capability will enable creators to enrich their videos with transitional scenes, enhancing storytelling. For instance, adding a backdrop of the New York City skyline at the start of a sightseeing video provides valuable context.
YouTube plans to integrate Veo into Dream Screen later this year, and all creations on Shorts will feature a watermark using DeepMind’s SynthID technology to identify them as AI-generated.
In addition to Veo, the company rolled out several new features, including “Jewels” and digital gifts that viewers can send during livestreams, similar to TikTok’s “Gifts.” Jewels aims to facilitate new interactions between viewers and creators, enhancing engagement during livestreams, with initial rollout focusing on vertical livestreams in the U.S.
YouTube has also extended its automatic dubbing tool to cover more languages, including French, Italian, Portuguese, and Spanish. The platform is testing a new feature called “expressive speech,” which captures a creator's tone, intonation, and ambient sounds for dubbed audio, fostering a more authentic viewer experience.
Furthermore, YouTube is expanding its Community hubs to a broader range of channels, encouraging interaction between creators and followers through shared posts and replies. The “hyping” feature, which allows users to show support for their favorite creators, is being introduced to more markets after its initial testing in Brazil, Turkey, and Taiwan; videos with the highest hype points will be featured on a special leaderboard.
Creators will also now have AI support for brainstorming video ideas within YouTube Studio, along with the ability to create AI-generated thumbnails and engage with followers using AI-enhanced comments.