Meta previewed new AI tools on Friday known as Film Gen that may create movies, edit them robotically, and layer on AI-generated sound for a cohesive video clip.
Film Gen works with written textual content prompts, a picture, or an current video as enter. There’s additionally an choice so as to add a private image so customers can see themselves within the video.
Associated: Meta Is Putting AI Images on Your Facebook and Instagram Feeds, With Personalized Pictures
After the AI works its magic and generates a video, a consumer can kind in a textual content immediate to create a customized audio soundtrack to play with the video.
A peek right into a video Meta Film Gen created from a picture. Credit score: Meta
Whereas Meta says Movie Gen’s high-definition videos are “the primary of its sort within the {industry},” when creating long videos at completely different facet ratios, it doesn’t suggest the AI is ideal. Proper now the AI can solely generate movies that last as long as 16 seconds — and it would not at all times get the project proper.
In a demonstration to the New York Occasions, Meta’s AI software made a mistake. Although it was capable of create a video of a canine in a park speaking right into a cellphone, the AI tousled by inserting a human hand across the cellphone as a substitute of a canine’s paw.
? Immediately we’re premiering Meta Film Gen: essentially the most superior media basis fashions to-date.
Developed by AI analysis groups at Meta, Film Gen delivers state-of-the-art outcomes throughout a spread of capabilities. We’re excited for the potential of this line of analysis to usher in… pic.twitter.com/NDOnyKOOyq
— AI at Meta (@AIatMeta) October 4, 2024
Chris Cox, chief product officer at Meta, stated in a Threads post that Film Gen is “industry-leading” in video high quality however that Meta is not ready to launch the instruments as a result of they’re too costly and the movies at present take too lengthy to generate.
Meta is sharing what it has proper now as a result of the outputs “are getting fairly spectacular,” Cox wrote.
Meta is not the primary to point out off a text-to-video AI generator software — ChatGPT-maker OpenAI did it in February with its text-to-video model Sora.
In July, OpenAI published multiple YouTube videos in partnership with artists and entrepreneurs displaying how Sora might create fantastical quick movies.
Associated: Mark Zuckerberg Does a Better Job Than His Rivals at Explaining AI