Until now, the term “AI art” has meant “still images”. No more. Meta’s Make-A-Video creates AI, where the company is combining AI art and tweening to create short, looping video GIFs.
Make-A-Video.studio is not yet publicly available. Instead, it’s being shown as what Meta itself can do with technology. And yes, while this is technically a video – in the sense that there’s more than a few AI art frames together – it’s probably still closer to a traditional GIF than anything else.
It doesn’t matter. What Make-A-Video accomplishes is threefold, given the demo on Meta’s website. First, the technology can take two related images—whether they’re a drop of water in flight or pictures of a galloping horse—and create the frames in between. More impressively, Make-A-Video seems to be able to take a still image and apply motion to it in an intelligent way, taking a still image of a boat, for example, and creating a short video of it moving through the waves.
Finally, Make-A-Video can put it all together. From a prompt, “a teddy bear painting a portrait”, Meta showed a small GIF of an animated teddy bear painting itself. This shows not only the ability to create AI art, but also to infer actions from it, as the company’s research work indicates.
Goal
“The Make-A-Video research builds on recent progress made in text-to-image technology designed to enable text-to-video generation,” explains Meta. “The system uses images with descriptions to know what the world is like and how it is often described. He also uses unlabeled videos to learn how the world moves. With this data, Make-A-Video allows you to bring your imagination to life by generating whimsical and unique videos with just a few words or lines of text.”
This likely means that Meta is training the algorithm on real captured video. What is not clear is how this video is being inserted. Facebook’s research work on the subject does not indicate how the video might be sourced in the future, and one has to wonder whether the anonymous video captured from Facebook could be used as a seed for future art.
This isn’t entirely new, at least conceptually. Animations like VQGAN+clip Turbo can take a text prompt and turn it into an animated video, but Meta’s work looks more sophisticated. It’s hard to say, though, until the model is released for an audience to play with.
However, this takes the art of AI to another dimension: that of movement. How long will it be before Midjourney and Stable Diffusion do the same on your PC?
