A few weeks ago, I suggested that the solution to our content moderation problem is federation. Instead of continuing a losing battle to improve how centralized tech platforms handle online content, I argued that we should push these decisions to the edge of the web. This way, servers can decide for themselves which other servers to connect to, while allowing their users to stay connected to global chats. Federation was the original design of the internet, and at the same time I believed that modern communications could function.
But even as I made the case for decentralization, I knew it would never be fully resolved. Decentralized systems, paradoxically, suffer from many concerns that can only be addressed through centralization. This means that unless we take proactive measures to prevent this from happening, the current shift in favor of decentralized solutions will meet the same fate as all previous efforts as the pendulum begins to swing back. So, what might these measures look like?
In a recent white paper, Varun Srinivasan argues for finding a balance between the over-centralized control and over-federation that characterizes our online interactions today – leaving out many of the features users expect in their online interactions. He calls this Goldilocks Zone “sufficient decentralization.”
According to Varun, only three things need to happen for a social media network to be sufficiently decentralized: (i) users need to be able to request a unique username; (ii) they must be able to post any message under that username; and (iii) they must be able to read that message on behalf of any user. The first of these has so far been extraordinarily difficult to put into place.
In a centralized network, the name registry is managed by the network operator. Names are usually assigned by accessibility, so “good names” are usually adopted by early adopters. This is how Twitter co-founder Jack Dorsey was able to secure the extremely popular @jack username. The problem is that the centralized platform operates under the control of the organization that controls it, and there is nothing to prevent you from denying access to your username or, worse, assigning it to someone else. Given how closely our social capital is tied to our online presence, this control over our online identity its absence is unacceptable for many.
Federated networks have different problems. Although users have more control over their usernames, they are loose networks of separately constructed instances and therefore have no way of identifying a given username as unique across the federated network. What this means is that in a decentralized network, your username is unique only on the server where it was created. Nothing prevents someone else from registering the same username on another server. This means that no one can have a unique username within ‘fediverse’.
For example, even though I’ve provided a very common username @rahul for myself on Mastodon, you’ll need to search for rahul@thinktanki.social to find me because I’ve registered my username on the thinktanki.social server. This is what separates me from the other Rahul who got lucky enough to register on the (more popular) mastodon.social server.
We need a decentralized name registry so that usernames can be uniquely identified in a federated social network. Until recently, this was believed to be impossible at scale. Varun suggests that we can change this by using smart contracts to build a decentralized name registry. Each new username can be added to the ones before them in the chain, which in turn serves as a common (yet decentralized) registry of names for all applications connected to the protocol. By doing so, users gain exclusive control over their usernames while taking advantage of a federated network.
This idea is built on a new fairly decentralized social network called Farcaster. By keeping the username on-chain and decentralizing the storage of posts in Farcaster Hub, users can have a modern social media experience with complete control over their online identity.
Farcaster is just one of a number of protocols and solutions built to offer a significantly different online experience than what we’re used to. Cumulatively, these platforms are called Web 3.0, which is not really different from the static web (Web 1.0), which only allows us to read the content uploaded to it, or the dynamic web (Web 2.0), where users can create and consume content. own it – envisioned as a decentralized and self-reinforcing network that takes control away from platforms and returns it to individual users.
If successful, these new protocols will shape a very different online future than we expect. As Balaji Srinivasan elaborates, when we can rely on blockchain, I can see us producing the first draft of history from the raw facts written on a distributed ledger. Once smart contracts are everywhere, our laws will be written directly into code, interpreted by impartial servers, and cryptographically enforced.
This is a completely different version of the internet than we use today. Whether it’s better or worse remains to be seen.
Get all Business News, Market News, Latest News Events and Latest News Updates on Live Mint. Download The Mint News app to get daily market updates.
Less