Meta, the corporate home of Facebook, just launched Make-A-Video, a new AI system that lets people turn text prompts into video clips, sort of like a video take on the controversial DALL-E and Imagen text-to-image generators.
“Generative AI research is pushing creative expression forward by giving people tools to quickly and easily create new content,” as the Meta AI blog notes. “With just a few words or lines of text, Make-A-Video can bring imagination to life and create one-of-a-kind videos full of vivid colors, characters, and landscapes. The system can also create videos from images or take existing videos and create new ones that are similar.”
Meta seems to be aware of recent controversies with AI-generated content, which has led to concerns about everything from copyright to deep fakes. The company is using publicly available datasets, openly sharing its research and results with the community utilizing a “responsible AI framework.” They’ll also be adding a watermark to all created videos noting they were created with AI. It’s currently not available to the public.
Currently, these generated clips are about five seconds long, and the examples Meta lists are pretty benign (“A teddy bear painting a portrait,” which is admittedly adorable). It can also be used to add motion to a still image or create variations on existing video.
MIT Technology Review suggests that the current level of tech is crude and has the “aesthetics of a trippy old home video,” but the technology itself does raise some ethical issues, as the tech requires immense power (limited its use to giant companies) and is tricky to “train.” The tech itself has the potential to create “unique harms that could be caused through generated video as opposed to still images,” says Henry Ajder, an expert on synthetic media.
Thanks for reading InsideHook. Sign up for our daily newsletter and be in the know.