Super excited to share what I’ve been working on: Odyssey!
We’re building Hollywood-grade, directable visual AI, to enable storytellers to create new, amazing movies, TV shows, and video games. We believe this technology is what comes after text-to-video.
Firstly, let’s talk about what has shaped our technology approach more than most: autonomous vehicles (AV).
Myself and Oliver Cameron, plus a large chunk of our team from Cruise, Wayve, Waymo, and NVIDIA, have spent years building AV technology.
Why is AV relevant?
1️⃣ AV had to push the SOTA in visual AI to ship.
2️⃣ AV scaled performance with real-world data.
3️⃣ AV necessitated scalable, physics-accurate simulation systems.
AVs are taught to navigate the world.
Odyssey is taught to generate worlds.
After pushing the SOTA in visual AI for many years, and watching all the progress in generative ML, we were confident that Hollywood-grade outputs could be generated by AI.
The question was then: what abstractions would make this technology useful for movie-making and more?
If I told you ChatGPT output text you couldn’t edit, you wouldn’t use it. But that’s how text-to-video works today, where the model output is a fully-baked video file that you can’t edit. If any of the scenery, characters, lighting, or motion is off, you have to start again.
Given our experience in AVs, our minds naturally went to 3D, and generating layers of 3D worlds. This is why we’re training models to generate…
🧊 High-quality geometry
🏔️ Photorealistic materials
🌄 Stunning lighting
🏃♀️ Controllable motion
With this approach, our users will still be able to generate video at the speed of AI, but have fine-tuned control over all the minutiae that matter.
We’re early on our journey, but the progress has been incredible, and we’re ready to work alongside Hollywood to shape this. Please get in touch!
I’m fortunate to work with some world-class folks to bring this to life. We’re now growing our R&D team for the right individuals. If you’re excited about working with the best visual AI team in the industry, check out our jobs section.
Lastly, a big thank you to our amazing supporters in GV (Google Ventures), Air Street Capital, DCVC, and our many wonderful angels.
https://odyssey.systems
Optimizing Business with AI and Data
1wThis is awesome. Just watched an interview on X with Peggy Johnson, CEO of Agility Robotics and quickly became a fan. She has both a pragmatic way about her but clearly has the big picture in mind. One thing she touted was how Digit (Agility's robot) benefits from business data specific the customer's facility that helps train AI models to learn new skills. She emphasized that the robots use "conventional methods" to navigate within a localized area instead of using AI. I noticed some QR codes on the floor which tells me that Digit requires a semi-structured environment to operate. While pragmatic, Ms. Johnson also has a firm grasp of the big picture, expressing that AI should be regulated because of their proximity to humans, and that humans should always be "top of mind" . GXO Logistics, Inc. appears to be in good hands with Agility. I have so many questions, though.. is an hourly wage paid to Agility? Are the robots rented? Purchased? Is the cost meter still running when idle? What about labor optimization? Can two robots work in the same area? Can one turn off when demand is low (thus reducing costs)? These are exciting times for warehouse labor optimization!!