Odyssey aims to bring Hollywood-grade AI video creation to the masses
Key differences from other text-to-video tools: Odyssey promises to give users full, fine-tuned control over every element in the generated scenes, unlike current text-to-video tools that output a non-editable video file:
- If the scenery, characters, lighting, or motion is off in current tools, the user has to start the generation process again from scratch.
- Odyssey overcomes this by training four separate generative models that allow fine-tuning of geometry, materials, lighting, and motion, enabling users to rapidly create videos exactly as envisioned.
Startup background and approach: Odyssey is a year-old startup founded by Oliver Cameron and Jeff Hawke, who previously worked on autonomous vehicle technologies:
- The visual AI techniques used in self-driving cars, such as accurate physics simulations and scaling with real-world data, inspired Odyssey’s approach.
- The company has raised $9 million in funding to build AI models that are harder to create than text-to-video tools but should result in a superior end product suitable for professional film and game production.
- Cameron emphasized that for AI to work in Hollywood, it must generate “glitch-free and mind-blowing visuals”, in contrast to the lower-quality AI-generated content currently prevalent.
Potential impact on the creative industries: If Odyssey delivers on its promise, it could significantly lower the barriers to creating high-quality, realistic video content and enable a wider range of storytellers to bring their visions to life in movies, TV, and games. However, key questions remain about the specific capabilities and limitations of the technology, as well as the timeline for making it accessible to creators.
Meet Odyssey — AI video that’s ‘fit for Hollywood’