In a recent exclusive interview, Meta, the parent company of Facebook and Instagram, unveiled its latest artificial intelligence model, Movie Gen. This cutting-edge tool aims to generate highly realistic video and audio clips in response to user prompts, positioning itself as a new contender alongside established media generation solutions from innovative startups like OpenAI and ElevenLabs.
Meta shared samples showcasing Movie Gen’s ability to create a diverse range of engaging content. These include videos featuring animals swimming and surfing, as well as clips where real images of individuals are transformed to depict them in various activities, such as painting.
Company representatives elaborated on the functionality of Movie Gen, explaining that it not only generates video content but also produces background music and sound effects that perfectly sync with the visuals. Users have the option to enhance existing videos with this technology. For example, one demonstration illustrated the model adding pompoms to a man’s hands as he ran through a desert, while another showcased a transformation of a dry parking lot into a lively scene with a skateboarder splashing through water.
The capabilities of Movie Gen allow for the production of videos lasting up to 16 seconds and audio clips extending to 45 seconds. Preliminary blind tests indicate that Movie Gen’s performance is competitive with offerings from well-known companies like Runway, OpenAI, ElevenLabs, and Kling.
This announcement comes at a pivotal moment for Hollywood, which is grappling with the integration of generative AI technology in film production. Earlier this year, OpenAI, backed by Microsoft, introduced Sora, demonstrating its ability to create feature film-like videos from text prompts. While many in the entertainment industry are eager to leverage such innovations to streamline creative processes, concerns regarding the ethical implications of using AI trained on copyrighted material without authorization are growing.
Lawmakers have also voiced concerns about the potential misuse of AI-generated content, especially deepfakes, in global elections, with notable examples emerging from the United States, Pakistan, India, and Indonesia.
Despite the excitement surrounding Movie Gen, Meta’s representatives indicated that the company is not planning to release this tool for open access to developers, a departure from the approach taken with its Llama series of large language models. They stressed the importance of evaluating the risks associated with each model individually but did not reveal specific assessments pertaining to Movie Gen.
Instead, Meta intends to work closely with the entertainment community and various content creators to explore the applications of Movie Gen. The company is also looking to integrate this new technology into its products over the coming year.
According to a blog post and research paper released by Meta, the development of Movie Gen involved a mix of licensed and publicly available datasets.
Meanwhile, OpenAI has been engaging with Hollywood executives and agents to discuss potential partnerships related to Sora, though no formal agreements have yet been established. The conversation around AI’s ethical implications has intensified, particularly following actress Scarlett Johansson’s allegations that OpenAI mimicked her voice for its chatbot without her consent.
In related news, Lions Gate Entertainment, renowned for its franchises like The Hunger Games and Twilight, recently announced a collaboration with AI startup Runway. The studio has provided Runway access to its film and television library for training an AI model, with the expectation that filmmakers can use this technology to enhance their creative projects.
