Although OpenAI made big news as they unveiled Sora on day five of their ‘12 days of shipmas,’ a competitor is also getting a lot of attention and rivaling the vanguard company and its exploration of what you can do with AI video generation.
As usual, Nathaniel Whittemore at the AI Daily Brief podcast is on top of it – producing an episode specifically to go over Google Veo 2 technology.
In the no-headlines show, exclusive to Veo 2’s debut, he sums up five big use cases for the new genAI, and shares some responses and feedback on something that’s likely to revolutionize video production in ways that will astound us.
“It’s definitely the control of physics that has people most excited,” he says, suggesting that Google “stole the limelight” with their video and image models.
Google Veo 2: Five Use Cases
Whittemore goes into five elements of how Google Veo 2 is going to turn over the apple cart.
The first one is in social media, where he references a lot of visual effects from another pioneer company, Pika, and its ‘cakify’ effect, as well as others like ‘squish’ and ‘crush.’
Another major use case is video advertising.
“The word that sums up the AI future is ‘more’,” he says. ‘We’re just going to have more of everything, and certainly we’re going to have more advertising.”
Enumerating the third use case as establishing shots, B roll, and drone footage, Whittemore talks about how the product of these AI systems will likely replace companies buying footage from stock libraries.
“Veo and Sora are both already incredibly adapt at these sorts of establishing of natural world shots,” he remarks.
Another big feature is storyboarding and brainstorming, where Whittemore suggests a timeline editor embedded in the product allows you to string together what you need to make a complex video.
Lastly, he suggests that Hollywood and pro filmmakers are going to embrace these technologies, but so are the amateur filmmakers waiting in the wings.
Democratization of Video Production
Here’s where I started to think about the realistic effect of these technologies.
All you have to do is go back and look at what happened to photography, songwriting, graphic design, the creation of poetry and prose – as AI made all of these things automatic, their value dropped, and we had to deal with an inundation of mediocre content into markets.
If that happens to video, the results will be profound.
“The future of video creation will be defined by AI as a co-creator, democratizing access to professional-grade tools and enabling anyone to turn ideas into compelling visual stories,” says Fei-Fei Li, Co-Director of the Stanford Institute for Human-Centered AI, and someone who we have quoted extensively in plumbing the depths of AI’s advent. “This shift will not only revolutionize how content is made, but also who gets to create it.”
Well said, in my view.
Veo 2: Interesting Demos
I watched the Google DeepMind sizzle reel of Veo 2 at work.
There were cute, fluffy animated creatures developed ex nihilo that make the productions of the 1980s look like a kid scribbling on a wall.
There were photorealistic renderings of male and female heads and bodies festooned with strange confetti-like colored goo.
There were landscapes both natural and fantastic, with skyscrapers draped in what looked like fondant.
And all of the physics seemed to be correct – early in the podcast, Whittemore referenced a ‘Turing test for video,’ and it seems that these technologies have passed that test with flying colors.
So we’ll have to wait and see what specific affects this has on multi-billion-dollar markets. Are film companies going to invest in those big casting and production projects that cost so much money, if they can just outsource the entire thing to a large language model? And when you watch video that’s created based on AI agents learning about real world physics, what is the effect on your mind? How does it inform your experience as a viewer?
Some of that is pretty heavy stuff, and you’ll probably want to sit with your feelings as you start to realize that what you’re seeing, either at home in streaming media or on the big movie screen, has been made with artificial intelligence.