Nick Simmons is the co-founder of Octu Ventures, a venture DAO that invests in Urbit infrastructure startups. He was previously the Head of Community and Special Ops at Tlon. You can find him on Urbit at ~simfur-ritwed.
.jpg)
What is the simplest way to describe Urbit?
Urbit is three things: it’s an operating system, an ID system, and a networking protocol. Let’s go through each.
Urbit OS is an open source full stack overlay operating system, which means it can run on top of any Unix system. You can run Urbit just fine on a MacBook Pro with a decent amount of memory, a Raspberry Pi, Android phones, AWS, Google Cloud, whatever. (It doesn’t yet run directly on metal, but there is an ambitious project underway to make that happen.) It’s a purely deterministic functional operating system, which means it’s just a list of events that your computer has done so you can roll that up and go put it somewhere else whenever you want. The OS is based on an interpreter called Nock and a high level programming language called Hoon that has some similarities to LISP.
Urbit ID is a namespace based on public key infrastructure. It’s basically a system of identifiers that are instantiated as Ethereum NFTs. So every Urbit name has public and private keys, and lives as an object on the Ethereum blockchain that you hold in an Ethereum wallet. This is important because it means that all of the “nodes” of the Urbit network can be owned by anyone who has the keys. So your Urbit is running and talking to other nodes and identifying them by this one blockchain ID signifier that you hold the private keys to, and then you identify yourself through public keys.
All these nodes communicate over the Urbit network, which is the networking protocol and is actually quite unusual. Peer-to-peer networks usually make people mostly think of things like BitTorrent, which is a revolutionary piece of infrastructure that works quite well. But think: what were you trying to do with BitTorrent? You are searching the network for the hashes that denote little bits of files and for most of those files there are multiple people sharing them so they’re not unique. This is what we call flat peer-to-peer. It’s optimizing for speed and availability of a given file that gets copied again and again. That’s great, but that doesn’t really help me if I’m running a peer-to-peer network and I want to talk to you, specifically. You’re not a bunch of identical hash-identified file fragments of a Harry Potter movie. You’re a person and I want to connect to the actual person. So for something like Urbit, I need to find where you are in the network topology relative to me. When you have millions of nodes in a peer-to-peer network, this requires a hierarchical search and that’s a built-in function for Urbit.
How is this different from the types of social networks and applications most of us use today?
Overall, Urbit is designed as an anti-fragile, user-owned network on both the individual and infrastructure sides. If your sponsoring star or your sponsoring Galaxy goes offline for whatever reason, you have the right to go find a new one, and that permission is on you. It’s very different from the state of the client-server Internet today where your, say, Twitter identity is in a database at a Twitter server farm. You don’t really have keys to it. You don’t have the ownership of your social graph. Ditto for eBay, Amazon, or any social media site. You’re trusting them with your data and your identity is living on their server. So the original dream with Urbit was to replace that system with trust based on your history of interactions with other users who are themselves verifiable.
A weekly dispatch featuring exclusive interviews with deep tech founders & a roundup of the most important deep tech news.
What are Urbit planets, stars, and galaxies?
There are three main levels of the Urbit network protocol, but keep in mind that they are all running the same OS. First, there are about 4 billion planets, which are the base level identity. Then there are about 65,000 stars that come bundled with planets and perform packet routing plus peer-discovery for their sponsored planets to planets that are sponsored by another star. So the first time that I’m trying to connect with you on Urbit, my planet will talk to my sponsoring star, which will go and find your star and then route my request to your planet.
Galaxies route packets between stars. The galaxy/star relationship is similar to the star/planet relationship in the sense that every star has a galaxy sponsor. There are 256 galaxies and they are the root nodes of the network. They’re essentially identities or participants in what you could call a DAO. They have a suite of Ethereum smart contracts by which they can vote and change certain things with the network architecture. So, for example, if you wanted to change the number of planets that are on the network, which is currently immutably baked in, the galaxies could vote and say okay, we’re going to push a new version. Galaxies are also seen as costly sources of truth for software updates. So if you’re running a large software concern with lots of trusted users on Urbit, owning a galaxy is probably going to be useful to you.
How does someone get a planet, star, or galaxy?
Tlon was the company that incubated the whole Urbit network and to support their work they sold galaxies to investors to help the project early on. In fact, many investment funds such as a16z and Founders Fund own galaxies. There’s also a Grants Program that’s still active where the Urbit Foundation will give out stars with the full complement of planets for technical and community work. There are also retail markets for planets, stars, and galaxies. All of these things–planets, stars, and galaxies–are NFTs and so they can be publicly or privately sold or traded. You can go on Open Sea and buy them. It’s still relatively early days for the network, but the address space is already quite distributed. In fact, I think there’s over 100 unique galaxy owners and that’s not strictly counting funds and individuals who own more than one galaxy. There are a lot of people with pretty diverse interests who own bits of Urbit
What is Tlon and what is its relationship to the Urbit Foundation?
Tlon was the first Urbit company. It started around 2013 and the plan had always been that once a certain level of technical maturity was reached, that there would be an Urbit Foundation. That is what happened and actually how I came to be involved, which was helping do some early research about how to set up incentives correctly for the Urbit Foundation around selecting board members and that kind of thing.
So at this point, the foundation has taken over most core development work. Tlon is now a product company. They have a hosting service and they also maintain the first social app called Groups. The foundation has responsibility for core development, some primitives, and some things like the PKI interface—Bridge—and now there are a whole bunch of other companies, too. And now there are also multiple funds investing in Urbit startups.
What is Octu Ventures?
This is a venture DAO I helped start. We’re one of several seed funds in different areas incubated by Hydra, a fund of funds from the MetaCartel Ventures community. We’re a seed fund and all of the members participate in deal flow, due diligence, and ideation about where the Urbit space is going, which I think is really crucial because it’s an extremely expansive space in terms of what could be built. We’re getting to a particular place right now where the core technical primitives are solid enough that it’s time to build on top of them. If you look at the classic “crossing the chasm” model, you’re looking for early commercial adopters who are scanning the landscape for that 10x improvement in how they’re going to make it out of the primordial soup. The reason I think it’s a great time to start an Urbit seed fund—and for what it’s worth we’re not the only one, there’s about five others that have popped up in the last six months—is that we are starting to see companies that started with ideas and a small team less than a year ago that are now building on Urbit.
Where are you finding Hoon devs to build Urbit native startups?
Fortunately there’s a large supply of Hoon devs from the Urbit Foundation’s developer education program. So there’s several hundred Hoon devs now and each one of those devs, once they get to an intermediate level, they’re able to build a business. One of the interesting things about Urbit is that a skilled Hoon dev is absurdly productive in terms of working code that can ship an app out to the network because the core of the code base is much simpler. The big intention of Urbit’s originator Curtis Yarvin was to cut the Gordian knot of that long history of computing tradition that optimizes for performance at the cost of simplicity. Urbit’s kernel can literally fit on a t-shirt whereas the Linux kernel has tens of millions of lines of code.
So each of those Hoon devs is a business, but also they can be the Hoon dev at an external startup that didn’t start out as an Urbit thing. They can come in and make it play well with Urbit. I’ve been seeing that explode in the last two or three months, whereas previously there would be a glimmer of interest from somebody doing a crypto thing, but Urbit was just not ready yet. Now it’s ready.
What kind of Urbit startups are you investing in at Octu?
I’m particularly bullish on startups that use Urbit’s infrastructure to accomplish something that has an existing customer demand and/or uses Urbit’s unique architecture to make something possible that wouldn’t have been possible otherwise. Specifically, I’m very interested in Urbit’s ID system, which I think is an undersung aspect of the whole thing.
What changed to take Urbit from this small and relatively insular community to an ecosystem that can handle venture investment?
There were some fundamental technical questions and debates that had to be resolved around how you store data in Urbit. But those basically got resolved in the last year and that was definitely a factor. I also think that it was just this snowballing effect of having more developers. That developer pool is itself a sign of technical maturity because when something breaks you have a bunch of people who are knowledgeable and can build a solution. There’s also been a lot of work to speed up the network. That work is ongoing, but that’s also played a very important part.
Why is now the right time for Urbit?
I think that blockchains and zero knowledge encryption are the way that we adapt to the fact that all our data is out there. It’s possible to have very granular data, but our society requires obscurity. Not just to avoid living in a panopticon or in totalitarianism, but also to allow new things to be built and grow and not be utterly path dependent. Over the last two years, I’ve seen a lot of people realize that when crypto systems get to a point where they’re big enough to be taken seriously as a threat to legacy systems, if they don’t have the ability to model reality and absorb shocks and actually give people reliable tools then you end up with the worst of both worlds. I think there’s been a realization among crypto people in general that we need more full-fledged tools, we need more of a holistic society, we need social coordination layers that incorporate all this. If you’re in the business of building an entire society, you’re going to have to replace everything. And Urbit is one of the few sufficiently holistic approaches.
As an investor, when you look to the future of Urbit, does it look more like Windows or Linux in terms of adoption by everyday people?
We’re on a standard 10 year fund cycle and I’m not going to say that toward the middle or end of that cycle that we won’t see companies that are going to come out of that and build something for the mass consumer. But not now. We really see the massive opportunity in terms of building primitives and deploying them to early adopters, including enterprise.
That said, some of the earliest enthusiastic non-technical adopters and cheerleaders for Urbit were writers. You look at people like Noah Kumin who’s doing the Mars Review of Books. It’s a fascinating experiment in that he’s doing the literary magazine and it’s in print and then there’s also a presence on Urbit. So there are already these self-propelled artists and thinkers on it, and that’s not quite the same as a mass consumer app, but they see that this thing is ready enough for them to use now.
In terms of when my mom or my grandma is going to go on Urbit remains to be seen. I think Urbit’s infrastructural and ownership-related qualities are substantially different from current tech and new paradigms of “mass adoption” via prosumer tool usage are very likely to arise. You might not need everyone’s mom to adopt it as “Urbit qua Urbit” for it to become the substrate that an awful lot of serious commerce, culture, and even society runs on top of.