Meta announced Thursday the release of two new generative video tools that might eventually be widely used by people on Facebook and Instagram.
One is a tool called Emu Video that generates videos that users describe using text. The other is an editing tool called Emu Edit with which users can further alter videos, again with text input. Both tools are based on the Emu foundation model, which Meta had already announced.
Emu Video is powered by two diffusion models and can generate 512×512 four-second long videos at 16 frames per second. The Emu model can “animate” user-provided images based on a text prompt, Meta says.
Mark Zuckerberg once heralded that video would become the main form of expression on Facebook. Now Meta seems to be saying that AI-generated content will be the next big thing on the platform.
Meta says the two tools are just “fundamental research” right now, but teases some pretty clear use cases in the near future. “We’re working on integrating these creative editing capabilities into video and photos for Facebook and Instagram,” Meta says in a blog post.
“Imagine generating your own animated stickers or clever GIFs on the fly to send in the group chat rather than having to search for the perfect media for your reply,” it adds.
As with most things Meta does, mainstreaming AI-generated video could be a blessing and a curse. It can be used to spread information and toxicity as easily as it can to get a cheap laugh on Facebook.