Former Self-Driving Tech Execs Launch AI Video Startup for Hollywood Films

Former executives from self-driving companies Voyage and Wayve have launched Odyssey, an AI startup building video generation tools for Hollywood.

Oliver Cameron, former CEO of Voyage, unveiled the company on X (Twitter), with plans to build visual AI solutions where “beautiful scenery, characters, lighting and motion can be both generated and directed.”

“We believe visual AI can be a new frontier for storytelling. But, we must hold it to high standards,” Cameron said. “Today, we're surrounded by low-quality AI-generated text and imagery. For AI to work for Hollywood, it must be capable of producing glitch-free and mind-blowing visuals.”

Cameron co-founded Odyssey with Jeff Hawke, Wayve’s founding researcher and technology vice president.

The startup has already raised $9 million in seed funding from Google Ventures, DCVC and Air Street Capital and brought in visual artists who have worked on blockbuster franchises including Marvel’s “Avengers,” “Dune” and “Godzilla.”

Odyssey has also secured backing from several big-name angel investors, including Elad Gil, Google AI lead Jeff Dean, former Cruise CEO Kyle Vogt and Garry Tan, Y Combinator’s CEO.

The startup’s mission is to create AI-generated visual content that’s of a higher standard than existing systems.

Related:Hollywood Visual Effects Provider Gets $200M to Develop AI-Powered CGI Tools

Cameron said there are too many “low-quality AI-generated text and image” solutions and that to make AI suitable for use in Hollywood movies, it needs to be better.

Models like Sora from OpenAI can generate short soundless videos from text prompts and still images. Researchers at Google DeepMind only uncovered how to add audio to AI-generated videos in June.

The Odyssey co-founder said more powerful video generation models are needed to make them suitable for use in movies.

“As a storyteller, you have little ability to direct your environment or characters, or to iterate on the finer details of your shot until it's just right,” Cameron said.

Odyssey has developed four generative AI models that create layers of visual detail, including object geometry, lighting and motion.

The model outputs are then combined to create the desired scene, allowing users to alter the video at the individual layer, rather than having to regenerate an entire scene.

Cameron said the model’s outputs can be integrated into existing pipelines in use in Hollywood “and beyond.” The startup’s website suggests they could also be applied to video game creation.

Odyssey is the latest in a growing list of companies hoping to provide AI tools to Hollywood and the wider creative industries.

Related:VisionWorks Chief AI Scientist Laurence Moroney on AI in Moviemaking

Former Google AI lead Laurence Moroney is now working with VisionWorks Studios to provide predictive modeling to Hollywood studios.

Meanwhile, DNEG Group, known for creating visual effects for blockbuster movies like "Oppenheimer," recently launched Brahma, a company developing AI-powered CGI tools to simplify the creation of visual effects.