Web2 social platforms have long been criticized for their bad censorship, lack of user control and ban on cross-posting. Decentralized social protocols have emerged to solve these problems.

Decentralized social networks provide users the autonomy to set up social networks and services in their own ways. They foster higher resistance to censorship, protect user privacy, and enhance data security and reliability. Serving as the infrastructure for the next generation of social platforms to be born, decentralized social networks will play an essential role in shaping the future of social media.

This article will explore the key features, technical design, censorship resistance levels, and prospects of current decentralized social networks by analyzing three representative projects: Nostr, Farcaster, and Subsocial.

Key Features

Decentralization and the spirit of Web3 are core to decentralized social networks. This is reflected in their three key features: 1) decentralized social networks can operate without relying on any centralized parties or controllers; 2) they use decentralized technologies to store and process data to ensure efficiency and higher security; 3) they uphold user freedom and privacy and have higher levels of censorship resistance.

At present, not one Web3 social network has entered the mainstream. The underlying reason for this is that the values of Web2 social platforms have been established. They have found comfortable boundaries and use channels that meet regulatory requirements. If you care about censorship resistance, you should try decentralized social platforms.

Sacrificing some degree of decentralization might make it easier for Web3 social products to be accepted by regulators. But it’s against the megatrend of censorship resistance. If we liken the model of Web2 social networks to public squares, then a more apt metaphor for Web3 social networks is sewers. The advantage of the sewer model is that influence converges in the network but it will not cause a monopoly, and it is difficult for nodes (users) to alienate.

We’ve shared in a previous article how the sewer theory could help us determine if a public chain is a good investment. In this article, we also emphasized the importance of privacy to public chains. We believe privacy should also be a core value of decentralized social products.

Technical Design

Nostr: a censorship-resistant protocol secured by public and private keys

Nostr adopts several methods to ensure censorship resistance. Firstly, it uses digital signatures to identify users. It means that, as long as a user has his private key, his account will never be deleted. Secondly, it consists of clients and relays where clients can fetch data from and publish data to relays of their choice. And anyone can run a relay and set up rules to censor certain types of content. It means that users can easily switch to other relays. And even if a user is banned by all relays, he can choose to set up a relay of his own to be able to send messages again.

The significance of Nostr is that it has created a simple and universal solution for messaging while ensuring the intactness of messages. Under its design, users will not lose their digital identity or followers if they were banned by a relay. And by publishing content to several random relays, users can effectively enjoy censorship resistance. Additionally, a relay can charge a fee for publication, which ensures that there will always be a server willing to take the user’s money in exchange for serving their posts.

Farcaster: a decentralized social network with on and off-chain components

Farcaster is a sufficiently decentralized social network that enables users to send text, images, audio, and videos securely and privately. Its hybrid architecture combines both on and off-chain components. The Layer1 blockchain manages user identities while a Layer2 network propagates updates between users.

For the on-chain part, Farcaster uses Ethereum-based registry contracts to issue unique ids to users and store the custody addresses of these ids. Users can sign a message with their IDs. Recipients can verify its authenticity by looking up the custody address in the registry contracts and verifying the signature. This offers full security and consistency of Ethereum.

For the off-chain part, Farcaster uses delta groups to synchronize the social network’s state among user-controlled servers called Farcaster Hubs. Users upload messages to Farcaster Hubs which share messages across the network and store all the messages of all users. Apps can connect to Hubs and subscribe to updates from the network.

This architecture allows for tamper-proof communication over untrusted networks without requiring trusted third parties. Users can bring various Web3 identities, such as ENS domains, into the Farcaster ecosystem seamlessly and incorporate them in a user-controlled way.

Subsocial: an open platform built on Polkadot and IPFS technology stack

Subsocial is an open platform for creating decentralized, censorship-resistant social networks and marketplaces, built with the Polkadot and IPFS technology stack. Subsocial is not a social network but a platform with all the necessary modules and web front-end that users can use to create their own decentralized social networks, such as the decentralized versions of Medium, Twitter, Reddit, or Instagram. On these social networks, users own their content and social graphs.

The decentralized social networks that users create on Subsocial are called spaces. Subsocial also gives the power of content review to each space; Subsocial does not censor or block content at the chain level, but each space can specify its own content rules. The community can remove certain content that involves scams and harmful information by electing reviewers, or decide together how to block content that is not welcomed by the community.

Censors, Censorship Resistance, and Incentives

Users’ desire for new social experiences has inspired the explosion of decentralized social protocols. And new social experiences are often accompanied by the emergence of new censorship paradigms. With Nostr’s launch of Damus attracting attention, the discussion on the censorship-resistant narrative of decentralized protocols has been reignited. We believe that decentralized social networking is a concept we should examine from both technological and social perspectives.

From the “Square Model” to the “Channel Model”​

Social networks in their traditional forms are deemed by many as like public squares. The scope of freedom of speech and restrictions on that freedom are dictated and censored by systems with centralized power distribution. And such architecture could easily lead to a monopoly of influence. Some may argue traditional censorships are effective in curbing scams, frauds, and deceits. However, there has yet to be a concrete study of its efficacy.

The charm of decentralized social networks is that they follow the channel model instead of the public squire model. The traditional way of social networking requires a user to “encounter” another in some form. However, in Web3, such an encounter is not necessary. Users only need to broadcast their messages to the network. It is more behavior oriented than interaction oriented. Like water in the sewer, tokens bring users together to interact with others without actually encountering them.

The channel or sewer model has emerged to be a new paradigm of social networking. Accordingly, the content review will be a public and fair process of collective decision-making instead of a process dictated by a few. Or else, users’ broadcasts will lose meaning. The token economics models of blockchains provide good references, drawing on which we can create a sustainable way of operation that balances censorship and anticensorship, incentives, and user activity for a better social networking environment.

In our opinion, a successful decentralized social network must have a robust economic incentive model to stimulate more vitality in the network. To achieve this, two principles need to be followed for optimal balance — voluntarism and superhumanism.

Voluntarism

Without economic incentives, decentralized social networks will be unrealistic utopias. However, we’ve seen the demises of all most all social projects that issue their own tokens. Does that mean there’s no future for social tokens? We think the crux of the problem is that these projects didn’t think through whether their token economics will stimulate voluntary behaviors among users. For decentralized social networks to gain real network effects, they need to empower voluntarism among users with token economics.

A token economy that follows voluntarism should try to capture value from users’ behaviors instead of users themselves. For instance, account transfer might be one of the most frequent user behaviors in Web3. That is users would want to transfer their accounts from one project to another without losing their data or social graphs. Also, decentralized social networks should have a diverse marketplace of censor services to ensure benign competition. Therefore, interoperability is critical. With a good interoperability framework, common grounds can be found among different communities on censorship.

Superhumanism

A second principle that decentralized social networks should follow is superhumanism. That is users have full control and ownership of their personal data and the freedom to choose from many social apps. This principle can be further reflected in governance, avatars, and algorithms. Users will enjoy freedom in all of these aspects and have superhuman experiences on Web3 social networks.

1. Freedom in Governance. Users will be able to express their ideas and opinions on these networks without interference or suppression from any third parties. There will not be unified rules. Anyone can enjoy freedom of speech. The incentive model of social networks will be censorship resistant, accommodating the self-defined rules of each user. There will be some form of meta-governance that ensures the use of financial instruments is well aligned with community governance. It will be super easy for communities to set up economic terms to balance the interests of their members. And leveraging the governance utility of tokens, all affairs within a community will be reflected monetarily and an open and free market will form that can cover “long-tail” affairs. In a previous article about governance in Web3, we have explained this idea in more detail:* The Innovation of Web3.0 Governance Mechanism From the Perspective of DAO’s Decentralization*.
2. Freedom in Avatars. Direct-to-avatar (D2A) is an emerging business model selling products directly to avatars (D2A) or digital identities. As it indicates, digital identities should not be just addresses, domains, or IDs, but highly personalized avatars. Like virtual goods facing difficulties transferring from one virtual world to another, the main challenge for avatars is also interoperability. Standardization of cross-platform assets will be conducive to the efficacy of protocol incentives as well as improving the overall censorship resistance level of the whole network. Therefore, decentralized social protocols with cross-chain and cross-platform features should be paid attention to.
3. Freedom in Algorithm. What a user can see and who he can interact with is determined by algorithms. Decentralized social protocols must support open algorithms to give users control of their algorithms, independent from any third parties. In these networks, users choose their algorithms to have desired experience and the network itself will also be more secure and flexible as a result.

Prospects

Nostr: Excellent censorship resistance

We believe Nostr is the most censorship-resistant decentralized social protocol at present in that clients can publish to multiple relays and users can change to other relays at will. In theory, users will not lose control of their identity no matter how frequently they change to other relays. And that is the largest value of decentralized social networking, i.e., realizing the mobility of identities through public and private key pairs.

Compared to other social protocols, Nostr is not blockchain-based. It is minimal and highly interoperable, which allows it to be used for building large-scale apps. It will be easier for developers to reach a consensus on open standards and iterate frequently to build advanced client apps. The downside is there is lacking an incentive layer in its architecture. Despite the fact that any user can set up a relay, there are currently only a few relays worldwide because there’re certain barriers to setting up a relay. Nostr may adopt one of the following ways to stimulate relay setup.

1. A fee-based model: When using Nostr, users can pay fees to relays for their service. This not only ensures revenue for relays but helps safeguard the security of Nostr.
2. A collateral-based model: users can pledge bitcoins to the Nostr network to get services from relays. The pledged digital currencies will be used to pay fees to relays, which in turn ensures censorship resistance.
3. A mining-based model: Nostr can incentivize relays to get mining rewards from publishing messages for users.

Farcaster: Backed by the network effect of Ethereum

As a decentralized social network built on Ethereum, Farcaster uses Ethereum smart contracts to realize its various social functionalities. Currently, there are over 30 apps built on Farcaster. Compared to traditional centralized social platforms, Farcaster has a higher level of censorship resistance as the apps will be run by smart contracts instead of controlled by any centralized entities. Its use of blockchain technology also ensures the security and privacy of user data.

However, Farcaster is still in its early stage. There will need a lot more development and testing for it to reach its ideal level of censorship resistance and usability. Additionally, the scalability issue of Ethereum will have an influence on how far it can go. Although Ethereum is constantly upgrading to solve this issue, it still poses a challenge. You can also refer to our previous research on Farcaster for more discussion: When Will Social Protocols Come of Age. In summary, Ethereum-based decentralized social protocols like Farcaster have great potential, but their level of censorship resistance is lower than Nostr.

Subsocial: The Future of Interoperability

We believe decentralized social protocols built on Polkadot and IPFS tech stacks like Subsocial have great potential too. Projects could achieve scalability and censorship resistance by leveraging the cross-chain capability of Polkadot and the decentralized storage technology of IPFs. Therefore, these tech stacks are quite suitable for building decentralized social protocols. Additionally, Subsocial is censor-resistant at the blockchain level. It stores users’ data and social graphs on the blockchain, ensuring the safety and privacy of such data. With these features, Subsocial’s prospect of success is optimistic. But it also has some limitations.

1. Technological limitations: Polkadot and IPFS technologies still have a higher barrier to entry compared to Ethereum. Therefore, using and building social apps on such protocols may not be the top priority for users and developers unfamiliar with these technologies. However, as Polkadot’s cross-chain functionality improves and Filecoin introduces the FVM, the situation will greatly improve.
2. Security issues: Data security is important in decentralized protocols. Due to the lack of centralized control in Subsocial, it is difficult to effectively secure its security, which is also a limitation that Subsocial faces.

Conclusion

In summary, from the perspective of resistance to censorship, Nostr is the best decentralized social protocol currently available. But it has not yet formed a viable business model. An important task for it would be to find a sustainable way forward while balancing the “voluntarism” in developer experiences and the “superhumanism” in end-user experiences. So far, Nostr is boosting the adoption of the Lightening Network. This integration is expected to improve user experience on the payment layer.

NFTs and the metaverse are also directions that decentralized social networks could consider. In this respect, Ethereum-based Farcaster has a natural advantage for gaining the network effect around its token. But when introducing tokens into a permissionless decentralized social network, it must be ensured that the token's value doesn’t depreciate and that the token can be used in various social use cases. Comparatively, Subsocial has a more sophisticated design in terms of token incentives.

Which protocol will likely win out? The three protocols have different technological designs and focus. Their strengths and disadvantages will vary depending on the use cases. In the future, they must continue to innovate and improve to meet users’ needs and increase censorship resistance.

In summary, an ideal decentralized social protocol should have strong resistance to censorship, high security, strong scalability, good user experience, good ecological construction, and fair governance mechanisms. Decentralized social protocols also face some challenges, such as scaling, user experience, and privacy protection, which may affect their practical applications and resistance to censorship. In addition, the resistance to censorship of decentralized social protocols also faces legal and policy challenges, as governments in some countries or regions may take measures to restrict or block these protocols.

Future users will gravitate towards social products with more “superhumanism” and “voluntarism”. Projects without centralized forms of governance like Nostr are freer in spirit. In other words, compared to innovations in performance and scalability, we are more optimistic about social projects with strong resistance to censorship and high security. While the Ethereum ecosystem and its governance model have their own strengths, we believe truly decentralized social networks built on the bitcoin blockchain or other anti-censorship, privacy blockchains will be more worthy of attention.