Introducing 6D.ai and Our Master Plan
What we’re building and why it matters.
Today is a great day! We are coming out of semi-stealth, and I can start sharing more about what we’re working on and more importantly, why we’re working on it. Anyone who’s been reading my blog posts over at Super Ventures will know that I’ve spent the last nine years of my life building products to unlock the potential of AR, and the ten years before that building and selling the software infrastructure that much of the wireline and mobile internet was built on.
So 6D.ai is the natural next step; a software company building the infrastructure that the AR apps of today and the future will be built on. Late last year my friend and Super Ventures partner Ori Inbar and I came up with the term “AR Cloud.” It describes the fact that if AR apps are going to virtually live outside the phone, then part of the “operating system” needs to be on-device, and part needs to live in the cloud, as the real world is too big for any one device. The AR Cloud enables the UX breakthroughs that lead to engaging AR apps. Persistent content, occlusion and real-time shared experiences all depend on sharing and synchronizing 3D computer vision data across devices, time and space.
For the AR Cloud to live up to its potential, some tough technical problems needed to be solved. Persistence and multi-player have been demonstrated several times over the past ten years (in fact my old startup Dekko solved them on iPad 2 in 2011) however the real problem to be solved is in making it reliable and easy to use. Occlusion could only be solved with power-hungry depth cameras with limited range and which didn’t work outside in sunlight. Semantic understanding of a scene was restricted to 2D image recognition. I’d learned from experience that the technology to solve these UX problems needed a combination of dense mapping of the world (every pixel, not just optical features), and 3D semantics to deliver a great UX. Then the technology needed to be able to run on mass-market smartphone devices without any add-ons, to reach a big enough support a business.
When I met my co-founder, Victor, he was in San Francisco seeking some advice on turning his research into a company. He had built an algorithm which could build a dense 3D model of the world in real-time on an iPad. I had been working on solving some of the big problems with AR app engagement over the past 9 years, and one of the biggest was the chicken-and-egg problem of human-scale (as opposed to vehicle scale) 3D spatial models being needed to allow digital and physical content to interact. No one would manually go and capture the data without a compelling app being there to justify the data capture. No one could build the app without the 3D model being there for it to interact with. Somehow the data needed to be captured at the same time as the app was run. Victor had the technology to solve this problem!
Soon after we met, we decided to work in parallel to get a company off the ground. After about a year of testing with customers and evolving the technology, we just needed a consumer grade 6DoF tracker to come to market — ARKit and ARCore answered those prayers (not exactly… we knew they were coming :-). So with Oxford University as our 3rd co-founder, we founded 6D.ai. Our mission was to solve the hardest computer vision software problems preventing developers from building engaging AR applications. To build API’s for the AR Cloud.
The immediate response to what we had created was astounding. Many incredible customers and partners reached out expressing interest in working with us. A dream team of talented team members is joining together on this shared mission. After careful consideration of our long term capital needs, we were fortunate enough to recruit Niko and General Catalyst to join our team on this mission.
The long game
I loved that from the beginning Elon Musk shared what Tesla was going to do over the long term. So, in the same spirit here is 6D.ai’s secret master plan:
- Build APIs to capture the 3D geometry of the world for AR apps
- Apply that 3D data to build semantic APIs, so apps can understand the world
- Partner & extend those APIs into the operating system for reality itself
All to give a superpower to people’s inherent creative genius.
For better or worse I jumped into AR about 9 years ago, and prior to this had the benefit of seeing from the inside both the wireline internet (at Ascend Communications) and wireless internet (at Openwave) revolutions. I believe that AR will usher in a similar transformation, from where I’m sitting now, everything feels the same. However it will take longer than we would like, will require significant capital, wisdom and some luck to thread the needle in the market. Building 6D.ai into the defining AR software company, by making AR developers successful, is the mission of the team we are building, and which I’m honored to contribute my time and energy.