The Guardian: Meta announces new AI model that can generate video with sound
https://www.theguardian.com/technology/2024/oct/04/meta-ai-video-generator-movie-gen
You are viewing a single comment's thread from:
The Guardian: Meta announces new AI model that can generate video with sound
https://www.theguardian.com/technology/2024/oct/04/meta-ai-video-generator-movie-gen
Meta Unveils Movie Gen: A Revolutionary AI Model Capable of Creating Realistic videos and Audio Clips
In a significant breakthrough in [artificial intelligence]
(https://inleo.io/@leoglossary/leoglossary-artificial-intelligence) Technology, Meta, the parent company of Facebook and Instagram, has announced the development of Movie Gen, a cutting-edge AI model that can generate realistic-seeming video and audio clips in response to user prompts.
This innovation positions Meta as a major player in the rapidly evolving field of generative AI, alongside leading startups like OpenAi and ElevenLabs.
Key features of Movie Gen
Movie Gen boasts impressive capabilities, including:
Realistic Video Generation: The model can create videos up to 16 seconds long, featuring realistic animations, such as animals swimming and surfing, or people performing actions like painting.
Audio Generation: Movie Gen can produce audio clips up to 45 seconds long, complete with background music and sound effects synchronized with the video content.
Video Editing: Users can edit existing videos using the model, allowing for seamless integration of new elements, such as inserting objects or changing environments.
Customization: Movie Gen can utilize users' photos to create personalized videos.
Performance and Comparison
Meta's benchmark tests demonstrate that Movie Gen performs favorably against offerings from prominent startups, including Runway, OpenAI, ElevenLabs, and Kling. Blind tests conducted by the company showed promising results, with Movie Gen's outputs often indistinguishable from real videos.
Applications and Partnerships
While Meta has no immediate plans to release Movie Gen for open use by developers, the company is collaborating with the entertainment community and content creators to explore potential applications. Movie Gen is expected to be integrated into Meta's products in 2025.
The entertainment industry has expressed interest in leveraging generative AI video technology to enhance and expedite film-making. However, concerns surrounding copyright infringement and the potential misuse of AI-generated deepfakes have sparked debate.
Concerns and Controversies
Lawmakers and industry professionals have raised concerns about the risks associated with AI-generated fakes, particularly in the context of elections and copyright infringement.
OpenAI's approach to training its Sora model has faced criticism, with actress Scarlett Johansson accusing the company of imitating her voice without permission.
The Future of AI-Generated Content
As Meta and other companies continue to push the boundaries of generative AI, the lines between reality and fiction are becoming increasingly blurred. The integration of Movie Gen into Meta's products and its potential applications in the entertainment industry will likely have far-reaching implications for the future of content creation.
About Movie Gen's Development
According to Meta's blog post and research paper, the company utilized a combination of licensed and publicly available datasets to build Movie Gen. This approach has sparked discussion about the ethics of training AI models on copyrighted materials without permission.
As the technology continues to evolve, it remains to be seen how Meta and other industry players will address concerns surrounding AI-generated content. One thing is certain, however: Movie Gen represents a significant leap forward in the field of generative AI, and its impact will be felt across various industries in the years to come.