Meta has just unwrapped a new potential tool for — or threat to — Hollywood.
The internet giant, the parent company of Facebook and Instagram, on Friday announced Movie Gen: a new AI tool that based on a text prompt can create realistic-looking videos clips (with synchronized AI-generated audio) that are up to 16 seconds in length at 16 frames per second. According to Meta, Movie Gen also can generate personalized AI videos using a single photo. Meta said it’s still testing Movie Gen before launching it publicly.
Mark Zuckerberg, Meta’s CEO and chairman, shared a video created by Movie Gen of himself doing leg presses — transported into different fake milieus, including one scene in which he’s dressed as a Roman centurion — on Instagram. “Every day is leg day with Meta’s new MovieGen AI model that can create and edit videos. Coming to Instagram next year,” Zuckerberg wrote.
A rep for Meta said, “While Movie Gen is purely AI research today, the models are essential to helping us build AI features that give you new ways to create and share highly engaging content on Facebook and Instagram.”
Meta shared multiple clips created using Movie Gen, including a surfing koala, a baby hippo swimming in a river (a nod to internet meme-celebrity Moo Deng), penguins dressed in Victorian outfits and a “woman DJ” spinning records. The company didn’t detail what sources of data it used for Movie Gen, saying only, “We’ve trained these models on a combination of licensed and publicly available data sets.”
“Whether a person is an aspiring filmmaker hoping to make it in Hollywood or a creator who enjoys making videos for their audience, we believe everyone should have access to tools that help enhance their creativity,” Meta said in a blog post announcing the tool.
At the same time, the company said, “While there are many exciting use cases for these foundation models, it’s important to note that generative AI isn’t a replacement for the work of artists and animators.”
Movie Gen’s audio generation model can take a video and optional text prompts and generate “high-quality and high-fidelity audio” up to 45 seconds, including ambient sound, sound effects and instrumental background music, all synced to the video content, according to Meta.
Meta outlines more technical detail for Movie Gen in a 92-page research paper, available at this link.
Here’s a 10-second clip of the swimming baby hippo generated by Movie Gen:
Other similar text-to-video systems include OpenAI’s Sora (which is not available publicly yet either). The advent of realistic-looking gen-AI video platforms has alarmed some in the media and entertainment biz, including Tyler Perry — who, citing Sora specifically, said he was suspending a planned $800 million expansion to his Atlanta studios.
Here’s Zuckerberg’s Instagram post with video created by Movie Gen: