About us
We're building an open, extensible hardware and software platform for running AI at home and at work.
We're exploring UIs that can support systems that are 1000x more intelligent than we have today.
Systems that can collaborate with a person on all levels of their life and work. Systems that anyone in the world can understand, extend, and improve.
Invention showcase
Substrate Strata
A easy-to-fabricate system for building expandable, hackable computers
Bridge
An always-on transcriber, translator, and assistant interface.
HTTP REFLECT
A new HTTP VERB to allow assistants to discover actions related to a specific URL.
Our approach
We believe AI is the future of both computing and the web.
We focus on technologies that operate on the individual and community-scale, and counterbalance today's global centralization and winner-take-all dynamics
We have a long term, positive-sum approach to building and investing in new technologies:
- Fund nimble teams of ambitious people for extended periods.
- Encourage those teams to invent new technologies that enable ambitious, ecosystem-scale goals.
- Set those technologies free. Open sourcing those technologies, and teach others how to use them.
- Actively inspire others to innovate and productize those technologies.
- Buy equity in promising commercial ventures deploying those technologies. Inviting others to do their own product development and market validation means a larger pie and greater leverage.
Frequently asked questions
Why run locally?
- Lots of local devices and sensors → HCI possibilities are awesome
- It's cheaper, more fun, and is better aligned with end users
- When users pay for their own equipment and electricity, there is no need to "monetize" them
Why a AI-specific computer?
- AI has extreme memory, compute, bandwidth requirements. The needs for space, bandwidth, cooling, and power are beyond that of phones, tablets, laptops and desktops.
- Enterprise-grade hardware can work, but it's expensive, loud, difficult to use, and hard to buy
- For things to keep moving, we need a common dev target for the entire ecosystem
Why open and extensible?
- Invites creativity and experimentation from others *and* our future selves
- Allows value accrue to the end users
- Ensures that such a powerful technology can't be captured by any single entity
Why an OS?
- Sub-second model loading requires tuning hardware, drivers, filesystems
- Easier to do whole-system optimizations (for example, per $ and per watt)
- Initial experience, upgrades, debugging are all easier (a smaller support matrix means higher quality)