Odyssey, a roughly year-old startup, says it's building a Hollywood-grade visual AI that can generate cinematic scenery, characters, and lighting.
What sets Odyssey apart from other text to video conversion tools is that it gives you complete, granular control over every element in your scene.
Odyssey co-founder Jeff Hawk argued that if ChatGPT generated text that users couldn't edit, they wouldn't use it.
“But that's how text-to-video conversion works today, and the output of the model is a fully baked video file that cannot be edited,” he wrote. X“If the scenery, the characters, the lighting or the movement is wrong, we have to start again.”
Creating a new type of model
I'm so excited to share what I'm working on: @odysseyml! We're building Hollywood-grade visual AI to empower storytellers to create amazing new movies, TV shows, and video games. We believe this technology is the next big thing after text-to-video. Learn more below. 👇 pic.twitter.com/XYr1Dcxo9rJuly 8, 2024
The way Odyssey overcomes this limitation is by training four generative models that allow users to fine-tune and control different layers of the video.
These models can generate high-quality geometry, photorealistic materials, compelling lighting, and controllable motion – if everything works as expected, you should be able to quickly generate the video you envision.
Oliver Cameron, co-founder and CEO of Odyssey, lamented that we are currently surrounded by “low-quality AI-generated text and images.” For AI to work in Hollywood, he said, it must produce “glitch-free, shocking footage.”
Cameron revealed that he has secured $9 million in funding from investors to fund the startup to bring these products to filmmakers. He said that their model is harder to build than text-to-video tools, but he believes it will result in a better end product that can be used in movies, TV shows and even video games.
Inspired by self-driving cars
Hawke said the AI video generator's approach, which could challenge the likes of Sora, Luma Labs Dream Machine and Runway, was actually largely born out of self-driving cars.
That's because for self-driving cars to hit the road, they needed cutting-edge techniques in visual AI to help scale their performance using real-world data. Self-driving cars also needed accurate physics-based simulations.
Cameron is co-founder of Voyage, a company that develops and deploys self-driving cars designed for seniors who are unable to drive.
Hawke was a founding researcher at Wave, a British company that tests self-driving cars in cities across the UK.
More articles from Tom's Guide
Back to MacBook Air
