OpenAI’s Disney Deal Brings Mickey Mouse & Marvel to Sora
▼ Summary
– Sora will generate short, user-prompted social videos using over 200 characters and assets from Disney, Marvel, Pixar, and Star Wars.
– These videos can be viewed and shared by fans and include animated characters, costumes, props, vehicles, and iconic environments.
– ChatGPT Images will create fully generated images from a few user words, using the same Disney-owned intellectual property.
– The image generation feature is designed to produce results in seconds.
– The agreement explicitly excludes the use of any talent likenesses or voices.
The recent partnership between OpenAI and Disney marks a significant step in bringing beloved characters and worlds into the realm of AI-generated content. This collaboration will see the integration of iconic Disney, Marvel, Pixar, and Star Wars intellectual property into OpenAI’s Sora video generation model and its image creation tools. The move allows fans to interact with these universes in a novel way, though it comes with specific creative boundaries.
Under this new agreement, Sora will be able to generate short, user-prompted social videos for fans to view and share. The system will have access to a vast library of more than 200 animated, masked, and creature characters from across the Disney portfolio. This includes not just the characters themselves, but also their associated costumes, props, vehicles, and famous environments. Imagine typing a simple prompt and receiving a brief video clip featuring your favorite heroes or settings.
In a parallel development, the image generation capabilities within ChatGPT will also be enhanced by this deal. Users will be able to type a few descriptive words and receive fully generated images in seconds, all drawing from the same rich well of Disney-owned characters and assets. This opens up new possibilities for fan art, creative projects, and personalized visual content.
It is crucial to note the limitations set by this agreement. The deal does not include any talent likenesses or voices. This means the AI models will not be trained on or authorized to replicate the actual faces or vocal performances of actors who portray live-action characters. The focus remains squarely on the animated and fictional elements, the costumes, creatures, and digital environments that define these stories. This distinction helps navigate the complex landscape of intellectual property and personal rights.
This strategic alliance represents a careful foray into generative AI for one of the world’s largest entertainment companies. By providing structured access to its digital assets, Disney can engage its audience with cutting-edge technology while maintaining control over its most valuable properties. For OpenAI, it offers a high-profile, legitimate use case for its models, showcasing their potential within established creative frameworks. The outcome will be a new form of fan-driven content, built from the building blocks of some of pop culture’s most enduring universes.
(Source: The Verge)
