(Credit: Google DeepMind)
AI was the star of the show at today's Google I/O keynote, where the company took the wraps off a new AI video generator, image creator, and music-making toolbox.
None are available to the general public yet, but they are part of an ongoing effort to encourage creatives to use AI tools.
"The creative potential for generative AI is immense and we can’t wait to see how people around the world will bring their ideas to life with our new models and tools," Google says.
The video generator, dubbed Veo, takes text-based prompts and turns them into 1080p resolution videos longer than 60 seconds. It is only available today to "select creators in private preview" and through a waitlist. Google plans to release it more widely in the future, as well as to YouTube Shorts as a new video-creation tool.
It's the eighth video generator Google has debuted in the past few years, following Generative Query Network (2018), DVD-GAN (2019), Imagen Video (2022), Phenaki (2022), WALT (2023), VideoPoet (2024), and Lumiere (2024).
But this one is the most capable yet, offering an "unprecedented level of creative control" and a higher quality output, Google says. It's also the first one Google has released since OpenAI grabbed headlines for its Sora video-to-text generator.
"Veo creates footage that’s consistent and coherent, so people, animals, and objects move realistically throughout shots," Google says. To assuage concerns over these tools replacing creative workers, Google enlisted Donald Glover for a video about the potential of the service to allow "everyone to tell their story."
Google also debuted Imagen 3, a new image generator. Like Veo, it is only available to select creators in private preview, though there's a waitlist. Compared to its predecessor, Imagen 3 is more adept at mimicking different styles of art, interpreting the intent of the text prompt, and rendering text in images, an area where AI image generators often stumble.
"This capability opens up possibilities for generating personalized birthday messages, title slides in presentations, and more," Google says.
Google also showed off Lyria, the AI music creation tool it launched in November 2023 that is helping artists "make entirely new songs that wouldn’t have been possible without these tools" and bring their ideas to life faster. Google enlisted musicians Wyclef Jean and Marc Rebillet to tout the benefits; both shared sample tracks they created with the new AI Music Sandbox.
Since AI-generated work cannot be copyrighted, the usual way creatives protect the value of their work, Google built its own solution. SynthId "embeds imperceptible digital watermarks into AI-generated images, audio, text, and video," including those made by its own products.
"We’re mindful about not only advancing the state of the art, but doing so responsibly," says Google. "So we’re taking measures to address the challenges raised by generative technologies and helping enable people and organizations to responsibly work with AI-generated content."