Video Generative AI: Why Is YouTube Already Using Robotic Intelligence? A Comment on Meta, SAG-AFTRA, and WGA
Movie Gen uses text inputs to automatically generate new videos, as well as edit existing footage or still images. The New York Times says the audio is matched with imagery with ambient noise, sound effects, and background music. The videos can be generated in either aspect ratio.
Movie Gen is able to make changes to an existing clip like adding an object into someones hands or changing the appearance of a surface. One example video from Meta depicts a woman wearing a virtual reality headset who is dressed in steampunk clothing.
Meta’s chief product officer, Chris Cox, writes on Threads that the company “[isn’t] ready to release this as a product anytime soon,” as it’s still expensive and generation time is too long.
AI image generators raised concerns about ownership and potentially harmful use cases, and AI video generators have only exacerbated them. Reports indicate that AI startup Runaway trained its video generator on thousands of scraped YouTube videos, something YouTube CEO Neal Mohan has already said would violate the platform’s terms of use. In a blog post, Meta says it trained Movie Gen on “a combination of licensed and publicly available datasets” but didn’t specify which.
Creatives like filmmakers, photographers, artists, writers, and actors also worry about how AI generators will affect their livelihoods, and AI has been a central part of several strikes, including the historic joint Hollywood strikes by the Screen Actors Guild – American Federation of Television and Radio Artists (SAG-AFTRA) and Writers Guild of America (WGA) last year.
Streaming Artificial Intelligence: The Movie Gen Experiment (META Connect) Releases 10 Second Clips and a Large Language Model
The company shared a number of 10-second clips from Movie Gen, including a Moo Deng-esque baby Hippo swimming around. While the tool is not yet available for use, this Movie Gen announcement comes shortly after its Meta Connect event, which showcased new and refreshed hardware and the latest version of its large language model, Llama 3.2.
Movie Gen can be used to generate audio bites. In the sample clips, an artificial intelligence man stand by a waterfall with splashes and sounds of a symphony, a snake slides along a jungle floor with horns, and a sports car roars and tires screech as it zips around a track.
Meta shared some further details about Movie Gen in a research paper released Friday. Movie Gen Video consists of 30 billion parameters, while Movie Gen Audio consists of 13 billion parameters. The largest variant of Llama 3.1 has 405 billion parameters, roughly equivalent to how capable a model is. Movie Gen can produce high-definition videos up to 16 seconds long, and Meta claims that it outperforms competitive models in overall video quality.
In order to demonstrate the Imagine Me feature, CEO Mark Zuckerberg showed users a photo of himself with gold chains on his hands and asked them to post it on Threads. The Movie Gen model makes it possible for a similar feature to be made into a video.
Considering Meta’s legacy as a social media company, it’s possible that tools powered by Movie Gen will start popping up, eventually, inside of Facebook, Instagram, and WhatsApp. The Veo model of video will be made available to creators inside of YouTube shorts later next year, according to a report in September.
While larger tech companies are still holding off on releasing video models to the public, you can experiment with artificial intelligence video tools from smaller, upcoming startups. Give Pikaffects a whirl if you’ve ever been curious what it would be like to see yourself cartoonishly crushed with a hydraulic press or suddenly melt in a puddle.
