- At I/O on Tuesday, Google announced a new AI-powered video generator called Veo.
- Google says Veo can generate videos longer than a minute and understands filmmaking terminology.
- This is a direct challenge to OpenAI's Sora tool, which was announced earlier this year.
Google is taking on OpenAI in a new field: AI-powered video creation.
Google's new Text-to-Video tool Veo can generate 1080p videos based on text, images, and video prompts, the company announced Tuesday at the Google I/O stage.
In its announcement, Google showed off some impressive clips created by Veo.
In one piece, a crocheted elephant ran through the savannah. Another photo showed an aerial view of a lighthouse on a cliff with waves crashing below. In the third clip, a person's face disappears in smoke.
Google says Veo can capture tones, simulate real-world physics, and even understand filmmaking terms like “time-lapse.”
“We're looking at features like storyboarding and generating longer scenes,” DeepMind CEO Demis Hassabis said on stage Tuesday. “Veo gives you unprecedented creative control.”
Google has announced that it has invited artists to participate in the Veo experiment.
“Over the past year, we've made incredible progress in improving the quality of generated media technology,” Eli Collins, Google's vice president of product management, and Doug Eck, senior research director, said in a blog post.
“We have been working closely with the creative community to explore how generative AI can best support the creative process, making sure our AI tools are as useful as possible at each stage.” The two are writing.
Veo invited a group of creators to explore the product, with actor and director Donald Glover and his creative studio Girga experimenting with the technology.
“Everyone is going to be a director, and everyone should be a director,” Glover said in a video shown at Google I/O. “Because at the heart of all of this is just storytelling. The more we can tell our stories to each other, the more we can understand each other.”
Google says the Veo waiting list is currently open.
The focus on creative collaboration comes amid criticism from entertainment industry veterans after OpenAI released its video generator Sora. They have previously been furious about AI tools as a danger to professions and the arts.
Film director Tyler Perry sees the technology as a threat, while director Tim Burton has previously said AI-made imitations of his films lack “humanity” and “soul.” was.
Still, some early testers thought the tool could facilitate the creative process, pointing out that Sora still requires human oversight.
Veo will be built within VideoFX, Google's recorder app, and Google says some features will be rolled out to YouTube Shorts and other Google products in the future.
To distinguish it as AI-generated, all videos created with Veo will advertise the new SynthID, an imperceptible digital watermark developed by Google.
Google says the company has been working on generative video for years, and Veo builds on previous ventures like Lumiere, VideoPoet, WALT, and Phenaki.
Just a few months after OpenAI made a splash with its unique AI-powered video creation tool, Veo has been revealed.
OpenAI released Sora in February, which includes eerily accurate scenes of people walking through a city and mammoths roaming in the snow.
Google appeared to take a shot at OpenAI's tools in Tuesday's announcement, noting that Veo can create clips longer than a minute. Sora's videos are limited to less than one minute of him.
Axel Springer, Business Insider's parent company, has a global deal that allows OpenAI to train models based on its media brands' reporting.
