Eli Collins, vice president of product management at Google DeepMind, first demoed the generative AI video tools to the company’s board in 2022. Despite the model’s ponderous speed, high maintenance costs, and sometimes erratic results, he says it was an eye-opening experience for them to see up-to-date video clips generated from a random prompt.
Now, just a few years later, Google has announced plans for a tool in the YouTube app that will allow anyone to generate video clips using artificial intelligence Veo model of this companyand directly publish them as part of YouTube Shorts. “Looking ahead to 2025, we’re going to allow users to create standalone videos and short films,” says Sarah Ali, senior director of product management at YouTube. “They’ll be able to generate six-second videos from an open text prompt.” Ali says the update could aid creators looking for footage to fill a video or trying to imagine something fantastical. She’s adamant that the Veo AI tool isn’t meant to replace creativity, but to augment it.
This isn’t the first time Google has introduced generative tools for YouTube, though this announcement will be the company’s most extensive video AI integration yet. Over the summer, Google launched an experimental tool called the Dream Screento generate AI backgrounds for videos. Before fully rolling out the generated clips next year, Google will update this AI green-screen tool with the Veo model in the next few months.
The emerging technology company has unveiled a number of AI-powered video models in recent years, including: Picture AND Lightbut it is trying to coalesce around a more unified vision with the Veo model. “Veo will be our model, by the way, in the future,” Collins says. “You shouldn’t expect us to have five more models.” Yes, Google will likely eventually release another video model, but it expects to focus on Veo in the near future.
Google faces competition from a number of startups developing their own generative text-to-video tools. OpenAI’s Sora is the most high-profile competitor, but the AI video model, announced earlier in 2024, is not yet publicly available and is reserved for a petite number of testers. As for tools that are widely available, AI startup Runway has released multiple versions of its video software, including a recent tool for adapting original videos into alternate-reality versions of clips.
YouTube’s announcement comes as generative AI tools become even more controversial for creators, who sometimes see the current wave of AI as stealing from their work and trying undermine the creative process. Ali doesn’t see generative AI tools coming between creators and the authenticity of their relationships with viewers. “It’s really about the audience and what they’re interested in — not necessarily the tools,” he says. “But if your audience is interested in how you did it, it’ll show up in the description.” Google plans to watermark every AI-generated video for YouTube Shorts SynthIDwhich embeds an unnoticeable tag to aid identify the video as synthetic, and also includes a “made using artificial intelligence” disclaimer in the description.
Hustle culture influencers are already trying game algorithm using a range of third-party tools to automate the inventive process and make money with minimal effort. Will next year’s Veo integration lead to a up-to-date wave of low-quality, spammy YouTube Shorts dominating people’s feeds? “I think our experience recommending the right content to the right audience lends itself to this world of large-scale AI because we’ve done it at such a massive scale,” Ali says. He also emphasizes that YouTube Standard Guidelines still apply regardless of the tool used to create the video.
AI art often has distinct aestheticswhich can be unsettling for video creators who value individuality and want their content to be unique. Collins hopes Google’s thumbprint won’t be evident on all AI video outputs. “I don’t want people to look at this and say, ‘Oh, this is a DeepMind model,’” he says. Getting the prompt to produce AI output that aligns with the creator’s vision is a primary goal, and avoiding an overt aesthetic for Veo is key to achieving broad adaptability.
“A big part of the journey is actually building something that’s useful to people and scalable and deployable,” Collins says. “This isn’t just a demo. This is being used in a real product.” He believes that putting generative AI tools directly into the YouTube app will be transformative for creators, as well as DeepMind. “We’ve never really built a product for creators,” he says. “And we’ve certainly never done it at this scale.”