In the fight to create the most cutting-edge AI technology in the world, Facebook parent company Meta Platforms Inc. upped the ante against competitors like OpenAI and Google with the release of a new AI tool that can create or modify movies in response to a straightforward text prompt.
Based on a written input, Movie Gen, a product of Meta, may generate a new video up to 16 seconds in length. These kinds of prompts can also be used to create audio for a video, modify an already-existing film, or even utilize a photo to make a personalized video with a real person in it.
Although Movie Gen is now only accessible to a small number of outside partners, including filmmakers, and select internal workers, the social media company—which also controls Instagram, WhatsApp, and Messenger—plans to integrate the product into its current apps at some point in the upcoming year.
The goal of Movie Gen is to inspire more people to produce or edit video posts, though executives are still debating the best way to carry out this integration, according to Connor Hayes, a vice president at Meta who specializes in generative AI products. We don’t currently have a clear product plan for it, but he stated it will be “fun to use, helpful for creators, good for overall engagement in the apps.”
The leader in social networking is only one of many large tech companies experimenting with artificial intelligence models targeted at producing films, which are more complex and costly to develop than algorithms that produce text. OpenAI, a company supported by Microsoft Corp., unveiled Sora, a video creation tool that can produce one-minute videos, earlier this year. However, the technology isn’t yet accessible to the general public. Another technology from Alphabet Inc.’s DeepMind is called Veo, and it was released earlier this year. Veo is a video generating tool.
The efficiency of the technology is one of the reasons Meta is delaying the release of Movie Gen. According to Hayes, it presently takes “tens of minutes” to create a video in response to a text query, which is too long for most consumers who will probably use this on their phone.
However, Hayes stated that Meta is also “sorting out a bunch of really important problems around safety and responsibility,” such as how to handle personalized movies, which prevents a user from making an offensive or derogatory video of someone else without that person’s permission. Prior to making the customisation feature widely accessible to people, it will likely be the most crucial issue to resolve.Although Taylor Swift and US President Joe Biden are two well-known figures who have been deceived by so-called “deepfakes” made with this kind of technology, Meta executives have stated that they are working on ways to “watermark” these creations so that viewers can recognize that they are artificial intelligence (AI) products.
AI developments are now a top goal for Meta as a whole, and CEO Mark Zuckerberg has frequently discussed AI’s role in driving user and revenue growth. Zuckerberg has given artificial intelligence (AI) short-term credit for enhancing the company’s content algorithms by displaying more pertinent articles and adverts to users. In the end, Zuckerberg stated that he thinks AI will be even more important in driving its applications and other cutting-edge wearables Meta is creating, such as smart glasses.