This is the introduction to the series Exploring Decentralized Social Networks that will cover the recent developments in decentralized and federated social media networks—specifically Mastodon, Farcaster, Urbit, Nostr, and Lens.
If you don’t care about setting the context and just want to get into the technical details, stay tuned for the first post that will explore identity across the different networks.
Mass discontent with the big social media platforms has been growing slowly over the past decade, but Elon’s purchase of Twitter and subsequent self-appointment to emperor, followed by a string of arbitrary decisions culminating in today’s decision to ban the promotion of other social media platforms has brought things from a simmer to a boil.
I don’t care too much about Elon as Twitter emperor or have opinions worth sharing about the positives or negatives of what he is doing. If you want that just go on Twitter and you’ll get an endless stream of it. What I do care about though is the immense amount of interest this has generated for decentralized alternatives.
It feels like a shift is underway and there’s a good chance that the social media landscape looks significantly different in a few years. In the past couple weeks I’ve been trying to get a better picture of what sorts of alternative landscapes are possible and I’m sharing my findings in this series.
The idea of decentralizing social media is not a new one. It’s well known that the early builders of the internet were idealists (with the irony of its foundations as a military project) who held decentralization as a core tenant. There has been much said on how we got from Web 1 to 2, so I won’t go into it here.
The important thing to understand is that it was never lack of imagination that led to the centralized internet we have today. Rather it was the simple fact that decentralized systems are harder to build, and neither the corporations creating the systems nor users of the systems cared to pay the additional cost. Centralized outcompeted decentralized alternatives with better user experiences.
Note: before blockchains identities were constraints by Zooko's Triangle–decentralized identifies could be only 2 of the 3: human-meaningful, secure, decentralized. More on this in Part 1.
Corporations were in fact happy to own the users data—turns out there’s a lot of money to being the intermediary of the world’s social graph. And users were happy to use these free services that provided superior experiences.
But alas, the piper must be paid. Whether they’re uncomfortable with the friendly collaboration between the government and social media moderators or worried about a billionaire with a collapsed OODA making seat-of-the-pants decisions or the algorithms that shape how we view the world in ways ever more powerful and beyond our understanding, there’s a new willingness on the part of users to bear the cost and move to new networks premised on a different set of social and technical ideas.
A decentralized social network is an ecosystem of protocols, computer networks, and applications that enable users to interact with each other without having to trust any particular third party.
There are many possible ways to build a DSN. Both technical and social design decisions will ultimately shape who is on the network and what kinds of interactions occur on it.
Regardless of the specifics, a social network can be said to be decentralized if it has the following features:
Sovereign Identity: anyone can create and maintain an account on the network
Permissionless Participation: any account can participate in (e.g. write to) the network at any time
Unrestricted Access: the record of any account’s participation in the network can be accessed at anytime
These were adapted from Varun Srinivasan’s essay Sufficient Decentralization for Social Networks, a piece I recommend you read. I’m exploring the concepts with you here and now, so take these with a grain of salt and not as decisive statements.
These features are made possible by a number of different technologies like cryptography, blockchains, peer-to-peer networking, and distributed backends.
I'll explore these features and technologies in detail in a four part series:
Part 1: Identity: how do accounts work in the network?
Part 2: Social Primitives: what kinds of social interactions are supported?
Part 3: Data Synchronization: how is data moved and stored within the network?
Part 4: Applications: how do users participate in the network?
Each of the concepts will be explored on its own and also by comparing the similarities, differences, advantages, and disadvantages in the approaches taken by Mastodon, Farcaster, Urbit, Nostr, and Lens.
Stay tuned for part 1 and in the meantime connect with me on Farcaster and let's discuss!
This is beautiful. Can’t wait for part 1!