Forward the Original Title:Breaking Down Walled Gardens: Farcaster and Decentralised Social Media
In today’s digital ecosystem, users find themselves confined within ‘walled gardens’ characterised as closed systems under the strict control of monolithic entities. These so-called digital platforms hold sway over user identities, data, and even the relationships between users, unilaterally setting rules for participation and developer interactions. Such centralised control erects formidable barriers against competitors and innovators in general. It enforces high user switching costs, is extractive and rent-seeking, and sometimes wields power comparable to governments.
Dominant digital platforms retain the sole capacity to modify, monitor, and selectively promote or censor information. They can access private messages, censor or elevate content, and even impersonate users. Despite assurances of transparency and user interest, the ultimate trust lies in the belief that those in power will not misuse private data or suppress dissenting voices. The spread of misinformation, driven by engagement-focused algorithms, erodes public trust. In a democratic society where the press is considered the ‘fifth estate’, entrusting speech control to centralised bodies, regardless of their intentions, poses a significant threat.
Social media has become a pivotal force in shaping contemporary narratives. It can influence public opinion, drive political discourse, and even impact global events. The authority wielded by these platforms has frequently sparked controversy as they face allegations of censorship, bias, and manipulation. This influence was evident during the “Arab Spring” when platforms like Facebook and Twitter played significant roles in mobilising protests and disseminating information, challenging authoritarian regimes across the Middle East and North Africa. Another example would be the “Rohingya crisis” in Myanmar.
One approach to enhancing the present state of social media is decentralisation, which entails converting centralised platforms into distributed networks comprising multiple independent nodes. Such a substantial shift necessitates a ground-up revamp of the existing architecture that is able to facilitate peer-to-peer interactions and collective content management. This change aims to disperse control and authority across a broader network while distributing value created by the platform, potentially leading to a more diverse and resilient range of social media interactions.
The decentralisation of social media is gaining traction as a response to increasing concerns over censorship, privacy, neutrality, user control, and malicious activity on centralised platforms. Decentralised social networks operate on independently run servers rather than on a centralised server owned by a single entity. The decentralised design grants greater autonomy and control to both users and independent developers, influencing the network’s functionality and the types of content allowed.
The concept can be likened to the analogy of the Cathedral versus Bazaar in software development. The Cathedral model is a centralised effort where a defined group of developers, or even a single one, is developing the software. This approach is more controlled and structured, reducing the scope of collaborative innovation. On the other hand, the Bazaar model is open and collaborative, with many people tinkering with the source code without central control. This method leads to rapid experimentation, innovation, and network stress-testing, as it allows for contributions from a diverse group of individuals.
Web3 brings forth many benefits for the realm of social media, tackling critical issues that are inherent in centralised systems:
Integral to every decentralised social network is a ‘protocol’, a shared language that ensures interoperability among different apps and services. These protocols can be likened to public infrastructure, similar to roads and sidewalks that facilitate movement between various destinations. Decentralised social media platforms are built on distributed architectures, where control and decision-making are shared among participants rather than being centralised in a single entity. Protocols are typically managed by core teams, often mission-driven non-profits, responsible for setting standards and ensuring a balanced, inclusive governance system.
Decentralised social media protocols are undertaking a variety of diverse approaches to data storage and identity handling:
The Web3 social landscape can be segmented into four layers, though this representation is not exhaustive:
Farcaster, for instance, is a decentralised social network that exemplifies the innovative use of these layers to create a more interconnected and user-empowered online social experience. In the following sections, we will explore Farcaster in greater detail.
Farcaster stands at the forefront of digital evolution, embodying a decentralised protocol crafted explicitly for creating and interlinking social applications. Its core mission is to forge a censorship-resistant environment, empowering users with absolute control over their data and audience connections. This approach signals a shift from traditional social media dynamics, offering a new realm of social autonomy and user empowerment.
The architecture of Farcaster is built on a decentralised network, allowing users to maintain a singular social graph across multiple applications. Envision a platform where diverse social apps, analogous to Twitter, Instagram, and Facebook, coexist harmoniously, linked by a single decentralised identity. This structure ensures that users retain their identity and network connections, even if individual apps impose restrictions. It’s a system designed to diminish the influence of centralised entities and return control to the users.
Farcaster’s network is permissionless and open-source, encouraging developers to engage and innovate by integrating with its APIs and other applications. This accessibility fosters an environment ripe for enhancing software functionalities and elevating user experiences.
At its core, Farcaster differs significantly from centralised apps like Twitter. Registration on Farcaster is based on a private-public key pair, specifically an Ethereum address. The on-chain aspect of Farcaster is primarily concerned with identity, where users mint a unique Farcaster ID (FID) that serves as their permanent identifier within the Farcaster ecosystem. While the FID is often referred to as an NFT, it’s important to clarify that it functions more as an identifier. Each account can have only one FID, as specified by the contract. This approach guarantees censorship resistance, as the FID is anchored on the Ethereum blockchain.
However, it’s important to clarify what exactly this mechanism achieves. While it guarantees that a message can be published to Farcaster, it does not inherently ensure that all people can read the message. Suppose a moderation policy is implemented on Farcaster that filters out messages from specific FIDs. This would still constitute censorship at the application layer. This highlights a nuanced challenge within decentralised platforms: while they can offer mechanisms for censorship resistance at the protocol level, the application layer can introduce its forms of content moderation and control.
Content, including the social graph, is stored off-chain within the Farcaster network, operated by entities known as Hubs. These Hubs function similarly to Ethereum nodes, with anyone having the capability to run a Hub. They ensure a consistent view of the network by syncing and exchanging messages.
The first application built on Farcaster is “Warpcast”, a Twitter-like platform. However, the potential to extend beyond this is far-reaching, with the possibility of other social apps, such as Instagram, YouTube, or Substack, leveraging Farcaster’s decentralised protocol. This system enables users to transfer their followers between apps, preventing apps from monopolising users’ social graphs and facilitating novel experiences that merge social activities with on-chain data.
Farcaster’s client-server relationship also sets it apart. Unlike Twitter, where a single client interacts with a centralised server, Farcaster allows for multiple servers, each offering diverse features. This unbundling of client and server reduces the risk of excessive power accumulation by any single entity, echoing the flexibility seen in exporting Gmail contacts to Outlook or moving assets between cryptocurrency exchanges.
The Farcaster protocol serves as the foundation, enabling developers to build competing clients on the same protocol. This approach is reminiscent of how different services, such as Substack and Mailchimp, are built on the Simple Mail Transfer Protocol (SMTP).
As a user-centric platform, Farcaster utilises the Ethereum blockchain to create a decentralised registry for user identities. Contrary to the need to generate a new wallet address, new users register a Farcaster ID (FID) with the Farcaster contract using their existing Ethereum address. This FID is linked to their username, and while it is unique and serves as a permanent identifier within the Farcaster ecosystem, it is not an NFT in the traditional sense as it is singular and non-transferable per account. This system minimises the need for blockchain interactions for everyday actions, improving the user experience by avoiding regular gas costs.
Farcaster’s innovative approach to social networking includes three core layers: the Identity Layer, Data Layer, and Application Layer. The Identity Layer, based on Ethereum, manages operations and authorisation. The Data Layer stores authorised information, while the Application Layer consumes this data.
The protocol addresses several key challenges of decentralised social networks:
Farcaster’s implementation of these solutions has led to the development of various applications, including Alphacaster, Discove, Jam, Opencast, Warpcast, and Yup. These applications leverage Farcaster’s protocol to offer diverse, user-centric social experiences similar to traditional platforms like Twitter but with enhanced user autonomy and network portability.
Hubs form a distributed network of servers essential for storing and validating Farcaster data, crucial for both reading and writing within the Farcaster ecosystem. They initiate operations by syncing with Farcaster contracts on the blockchain to recognise every user’s account and keys. The process for a Farcaster message includes its creation, signing, uploading to a Hub, validation, and then distribution via gossip to peer Hubs.
Validation checks for a valid signature and compliance with protocol specifications. Storage conflicts, like duplicates or limit exceedances, are resolved using CRDTs for deterministic outcomes. Hubs distribute messages using the gossipsub protocol and ensure data integrity through periodic diff syncs with peers, achieving strong eventual consistency even after disconnections. However, messages may arrive out of sequence.
There is no need for Hubs to reach consensus among themselves on the content, as the on-chain components ensure the integrity and ownership of messages through cryptographic signatures. In cases where two Hubs have conflicting messages, the on-chain data can be used to verify the authenticity and origin of the messages, resolving conflicts based on the immutable records on the blockchain. Hubs also employ peer scoring to assess behaviour, maintaining network integrity by sidelining peers who fail to meet standards.
Farcaster recently introduced a major feature called Farcaster Frames. A Frame is a standard designed to enhance user interaction by seamlessly integrating external content within the app’s interface. It turns casts into interactive mini-apps within Farcaster. Frames blend static or animated visuals with interactive elements. This allows users to interact with various external links and applications without exiting the app. These interactive buttons, upon activation, dispatch a POST request carrying a signed payload, facilitating a range of dynamic user engagements. Central to Frames’ operation is the EdDSA authentication system provided by Farcaster, ensuring a seamless and secure user experience across various platforms without necessitating application switches, thereby circumventing potential mobile operating system issues related to deep link redirects.
The introduction of Frames has led to a significant increase in Farcaster’s Daily Active Users (DAU), with developers rapidly adopting the platform to integrate native crypto functionalities like NFT minting and gaming. This surge in adoption shows Frames’ potential to bridge the gap between separate, siloed experiences, offering a unified and immersive user experience. Farcaster is actively developing enhancements for Frames, focusing on features such as text input, on-chain transactions, and improved privacy measures.
Farcaster, a nascent platform in the realm of decentralised social networks, is fostering a vibrant ecosystem of diverse applications, each catering to different facets of digital interaction and online community building. Below, we outline a couple of the exciting applications that are utilising Farcaster in unique ways:
Warpcast is a flagship application in the Farcaster ecosystem, mirroring the user interface of traditional social networking sites like Twitter. This Web3 app, accessible via mobile and web browsers, allows users to share posts (casts), engage with others, flaunt their NFT collections, and seamlessly integrate their on-chain activities into their social feeds.
Paragraph reimagines publishing in the decentralised world. Drawing inspiration from Substack in the Web2 space, it enhances user experience with extensive customisation options, automated email workflows, in-depth analytics, and collaborative tools for teams. Its unique offerings include minting, NFT memberships, token-gating, and direct integration with the Farcaster social graph, unlocking new avenues for content monetisation and audience engagement. When users sign up with an Ethereum wallet, Paragraph immediately taps into their Farcaster network to recommend newsletters from their connections. It allows users to subscribe to their followers’ newsletters, view a feed of their followers’ long-form content, and access trending posts shared within the Farcaster community. Paragraph also aggregates discussions from Farcaster directly onto the relevant newsletter posts, centralising conversations and enhancing engagement. The Paragraph team recently integrated Farcaster Frames to streamline the newsletter experience, offering users a one-click subscription to newsletters, and users can read the post directly from their Farcaster feed.
Kiwi News has emerged as a crypto-focused media dApp where community participation is centred around an NFT pass. Users can share and vote on a plethora of content, such as podcasts, news articles, and videos. Kiwi News, available as a web extension and app, offers an ad-free experience devoid of social noise, emphasising community-curated content.
Wield is a wallet that offers a unique club membership model, enabling users to earn rewards through crypto transactions. It leverages the Farcaster Layer 2, Cast, for efficient message submission to a federated, open-source Hub. This approach eliminates on-chain storage fees and Farcaster ID fee requirements. Farquest and Cast are applications built on the Wield platform that enhance the Farcaster protocol experience. Cast is a prominent Farcaster client with roughly 15,000 daily active users, while Farquest gamifies the protocol, allowing users to explore it in an RPG style and earn rewards.
The Farcaster community has developed a series of utility-focused applications, including:
The graph provides a view of user registrations for the protocol. Initially characterised by a modest uptick in registrations, the graph delineates a marked escalation post-July 2023 as it marked a significant pivot to a permissionless paid subscription model. The narrative above the graph is imbued with a sense of expectancy for 2024, reflecting the bullish stance on the future trajectory of user engagement with the protocol.
These graphs offer a granular view of the dynamics of interaction within the Farcaster community, highlighting content-sharing patterns and the corresponding engagement levels across different types of casts. The first graph tracks three key metrics: links, reactions, and casts. Peaks in the graph suggest sporadic surges in activity, with notable spikes, particularly in the reactions component, which indicate viral content or events prompting increased user interaction.
The total reactions of casts after 24h graph provides a breakdown of the volume of reactions received by casts within 24 hours of posting. It’s segmented into reaction ranges, such as 0, 1-5, 6-10, all the way to over 100 reactions, showing the distribution of engagement across posts.
The data points from this dashboard paint a clear picture of a platform on the rise, demonstrating substantial and growing engagement. The total of 2.8 million casts by 306,000 users signifies a robust and active user base. The weekly and monthly active casters, at 49,000 and 53,000, respectively, further highlight this healthy recurring engagement and are indicative of a platform that users return to regularly.
The user growth rate of 11% and the high Gini Index of 90 suggest a concentrated level of activity among a smaller subset of users. This could be interpreted as a core, highly engaged community driving the platform’s interaction, a common trait in emerging platforms.
The heatmap’s representation of weekly cast activity offers valuable insights into user behaviour patterns, showcasing the times when engagement is most intense. This data could be instrumental in targeting peak activity periods to launch new initiatives or features.
The ‘Users According to Casts’ pie chart and the ‘Continued Engagement’ metric underscore the platform’s success in attracting and retaining users over significant periods. The segments showing engagement for one week, two weeks, one month, and over three months confirm a sustained interest, a crucial factor for long-term growth.
All these statistics collectively indicate that Farcaster is on the right path when it comes to building and expanding its user base.
Farcaster’s ascent is contoured by notable challenges and adoption barriers, particularly regarding user experience and the broader perception of decentralisation. This task is complicated by the existence of multiple clients, each with its own slightly tweaked UI, which is a crucial factor for achieving wider adoption. This technological chasm potentially deters mainstream users, who may value interface familiarity and content richness over the nuances of the underlying technology. A pertinent debate in this context is the public’s valuation of decentralisation and censorship resistance. While appealing to a privacy-conscious demographic, it remains unclear if the average user prioritises these aspects over convenience.
Furthermore, the discourse on whether absolute decentralisation is the end goal or if a semi-centralised model of moderation, akin to recent evolutions in platforms like Wikipedia, might be more palatable, is increasingly relevant. Farcaster, for instance, adopts a sufficiently decentralised approach to reap the benefits of decentralisation while retaining the ability to scale and combat spam. This hybrid approach offers a balanced juxtaposition of user autonomy with responsible content governance.
As Farcaster scales, maintaining the integrity of its user base and community culture presents another hurdle. The platform’s ethos of fostering a friendly and innovative digital milieu must withstand the pressures of growth, ensuring that expansion doesn’t erode these foundational values. Farcaster also faces stiff competition from not just entrenched Web2 entities but also other protocols building on SocialFi in Web3, making market penetration and user conversion a formidable task.
Overall, decentralised social media platforms, such as Farcaster, Mastodon, Minds, and Lens, offer various advantages over traditional centralised platforms, including user control over data, censorship resistance, and increased privacy. However, they face significant challenges in gaining widespread adoption, such as:
As the application layer on top of decentralised social media protocols increasingly gains traction, the influence that large applications can amass over the underlying protocol becomes an essential but often overlooked topic. Large applications can significantly steer the protocol’s development decisions, much like Gmail’s local policy changes can impact all other email providers, even within a decentralised email protocol. For a decentralised protocol to remain truly democratic and resistant to the influence of large entities, it must either have mechanisms to enforce rules against a large entity or support a competitive environment with a multitude of players that prevent any single one from forcing changes unilaterally.
As we peer into the future of decentralised social media, we stand on the brink of groundbreaking shifts poised to fundamentally transform how we interact online. These advancements herald a new era where our virtual experiences are not only more engaging and personalised but also crafted within a framework that champions user autonomy and collective governance.
Decentralised social media protocols will enable a wide array of secondary innovations. Exciting areas that we are looking into are games and commerce. Prospects include the development of games that leverage user engagement and wallet activity. These could echo the early days of Facebook gaming but without the notorious spam issues. The commerce aspect is equally promising, with opportunities to purchase NFTs and access specialised services directly through Farcaster. Importantly, these developments on Farcaster come without the risk of situations like Zynga vs. Facebook, where complementary solutions built on top of centralised platforms can suddenly turn into competitors, highlighting the platform risk associated with centralised platforms.
Although decentralised content moderation also requires a delicate balance between freedom of speech and preventing abuse, the difficulty lies in distinguishing between acceptable and unacceptable content, a task complicated by cultural differences and subjective interpretations. While no universally agreed standard of acceptability exists, democratic processes, such as those employed by Wikipedia editors or federated social networks, offer models for decentralised content moderation that are not controlled by private entities. The future of decentralised social media may involve a democratic approach to content moderation, where users decide what is acceptable through debate and deliberation, supported by technical tools and automation. This governance should be based on ethical principles, democratic control, and accountability, moving beyond simplistic censorship resistance or dictatorial administrative powers. Currently, these functions are often bundled together in a single entity, leading to issues of bias and lack of neutrality.
The intersection of social media and artificial intelligence (AI) has been an active topic of discussion primarily due to how AI algorithms utilise user data to target ads and content. Twitter and other social networks have leveraged AI/ML models to analyse vast amounts of personal data, tailoring user experience and personalising content and advertisements. This raises concerns over privacy and data ownership as these models are trained on the users’ data, yet they don’t belong to them. Decentralised social protocols introduce a fundamentally different dynamic where users own and control their social graphs. Users can “mint” posts and content, build their own “on-chain content collections”, and support content creators through microtransactions. In the context of recent developments in AI, decentralised social models offer a unique advantage for users to train their own AI models on their own data. As Fred Wilson discussed in his recent article, users’ on-chain collections can serve as personalised data sets for training these AI models. This allows for highly personalised AI experiences, as the models are trained on data directly relevant to the user, and it respects user privacy because users can decide when and how their data is used.
A more modular approach, where different aspects of the platform are managed independently, could help address these issues. This would involve separating the baseline protocol, which should be neutral, from the moderation layer, which could be tailored to the needs and preferences of different user groups.
In the broader context of decentralised social media, this modular approach could lead to more neutral “public town squares” while still allowing for effective moderation. This would make the end-user experience more valuable, meaningful, fun, and useful, as well as “appropriate” by filtering out harmful actors and posts.
Farcaster’s expansion into broader functionalities is notable. This includes implementing verification mechanisms for genuine user giveaways and allowing Farcaster accounts to be used for logging into external applications, thus enhancing ecosystem connectivity. Despite being in its infancy, Farcaster is showing immense potential and generating excitement reminiscent of the early days of platforms like Clubhouse. Its focus on building a Web3-centric, creator-focused culture is evident. As Farcaster continues to grow, the questions about the scalability of its quality user base will become increasingly important. The platform’s commitment to fostering friendly, optimistic, and innovative online spaces distinguishes it in the digital landscape. Despite facing market challenges, its strategic vision and unique approach to decentralised social networking inspire optimism and anticipation for its future developments.
Web2 platforms, having matured over time, are bolstered by strong network effects but simultaneously challenged by numerous deeply rooted design problems. In contrast, Web3 social media platforms are nascent and often face criticism for their less polished user interfaces. Yet, they embody the promise of a better and more equitable future, aiming to dismantle the ‘walled gardens’ of traditional social media. As this new era of social media unfolds, it will be shaped by the collective efforts of its users and the ethical frameworks they establish, striving to create a more transparent, accountable, and inclusive digital ecosystem for all without the redundancy of past models.
We’re incredibly excited to see the emergence of new models and primitives in this sector. If you’re in the process of creating or planning to develop something innovative in the new era of Social, please don’t hesitate to get in touch with us.
We would like to express our sincere thanks to Anthony from Seed Club Ventures, JC & Nico at Wield and Katie from Archetype for their invaluable feedback on this report.
Share
Content
Forward the Original Title:Breaking Down Walled Gardens: Farcaster and Decentralised Social Media
In today’s digital ecosystem, users find themselves confined within ‘walled gardens’ characterised as closed systems under the strict control of monolithic entities. These so-called digital platforms hold sway over user identities, data, and even the relationships between users, unilaterally setting rules for participation and developer interactions. Such centralised control erects formidable barriers against competitors and innovators in general. It enforces high user switching costs, is extractive and rent-seeking, and sometimes wields power comparable to governments.
Dominant digital platforms retain the sole capacity to modify, monitor, and selectively promote or censor information. They can access private messages, censor or elevate content, and even impersonate users. Despite assurances of transparency and user interest, the ultimate trust lies in the belief that those in power will not misuse private data or suppress dissenting voices. The spread of misinformation, driven by engagement-focused algorithms, erodes public trust. In a democratic society where the press is considered the ‘fifth estate’, entrusting speech control to centralised bodies, regardless of their intentions, poses a significant threat.
Social media has become a pivotal force in shaping contemporary narratives. It can influence public opinion, drive political discourse, and even impact global events. The authority wielded by these platforms has frequently sparked controversy as they face allegations of censorship, bias, and manipulation. This influence was evident during the “Arab Spring” when platforms like Facebook and Twitter played significant roles in mobilising protests and disseminating information, challenging authoritarian regimes across the Middle East and North Africa. Another example would be the “Rohingya crisis” in Myanmar.
One approach to enhancing the present state of social media is decentralisation, which entails converting centralised platforms into distributed networks comprising multiple independent nodes. Such a substantial shift necessitates a ground-up revamp of the existing architecture that is able to facilitate peer-to-peer interactions and collective content management. This change aims to disperse control and authority across a broader network while distributing value created by the platform, potentially leading to a more diverse and resilient range of social media interactions.
The decentralisation of social media is gaining traction as a response to increasing concerns over censorship, privacy, neutrality, user control, and malicious activity on centralised platforms. Decentralised social networks operate on independently run servers rather than on a centralised server owned by a single entity. The decentralised design grants greater autonomy and control to both users and independent developers, influencing the network’s functionality and the types of content allowed.
The concept can be likened to the analogy of the Cathedral versus Bazaar in software development. The Cathedral model is a centralised effort where a defined group of developers, or even a single one, is developing the software. This approach is more controlled and structured, reducing the scope of collaborative innovation. On the other hand, the Bazaar model is open and collaborative, with many people tinkering with the source code without central control. This method leads to rapid experimentation, innovation, and network stress-testing, as it allows for contributions from a diverse group of individuals.
Web3 brings forth many benefits for the realm of social media, tackling critical issues that are inherent in centralised systems:
Integral to every decentralised social network is a ‘protocol’, a shared language that ensures interoperability among different apps and services. These protocols can be likened to public infrastructure, similar to roads and sidewalks that facilitate movement between various destinations. Decentralised social media platforms are built on distributed architectures, where control and decision-making are shared among participants rather than being centralised in a single entity. Protocols are typically managed by core teams, often mission-driven non-profits, responsible for setting standards and ensuring a balanced, inclusive governance system.
Decentralised social media protocols are undertaking a variety of diverse approaches to data storage and identity handling:
The Web3 social landscape can be segmented into four layers, though this representation is not exhaustive:
Farcaster, for instance, is a decentralised social network that exemplifies the innovative use of these layers to create a more interconnected and user-empowered online social experience. In the following sections, we will explore Farcaster in greater detail.
Farcaster stands at the forefront of digital evolution, embodying a decentralised protocol crafted explicitly for creating and interlinking social applications. Its core mission is to forge a censorship-resistant environment, empowering users with absolute control over their data and audience connections. This approach signals a shift from traditional social media dynamics, offering a new realm of social autonomy and user empowerment.
The architecture of Farcaster is built on a decentralised network, allowing users to maintain a singular social graph across multiple applications. Envision a platform where diverse social apps, analogous to Twitter, Instagram, and Facebook, coexist harmoniously, linked by a single decentralised identity. This structure ensures that users retain their identity and network connections, even if individual apps impose restrictions. It’s a system designed to diminish the influence of centralised entities and return control to the users.
Farcaster’s network is permissionless and open-source, encouraging developers to engage and innovate by integrating with its APIs and other applications. This accessibility fosters an environment ripe for enhancing software functionalities and elevating user experiences.
At its core, Farcaster differs significantly from centralised apps like Twitter. Registration on Farcaster is based on a private-public key pair, specifically an Ethereum address. The on-chain aspect of Farcaster is primarily concerned with identity, where users mint a unique Farcaster ID (FID) that serves as their permanent identifier within the Farcaster ecosystem. While the FID is often referred to as an NFT, it’s important to clarify that it functions more as an identifier. Each account can have only one FID, as specified by the contract. This approach guarantees censorship resistance, as the FID is anchored on the Ethereum blockchain.
However, it’s important to clarify what exactly this mechanism achieves. While it guarantees that a message can be published to Farcaster, it does not inherently ensure that all people can read the message. Suppose a moderation policy is implemented on Farcaster that filters out messages from specific FIDs. This would still constitute censorship at the application layer. This highlights a nuanced challenge within decentralised platforms: while they can offer mechanisms for censorship resistance at the protocol level, the application layer can introduce its forms of content moderation and control.
Content, including the social graph, is stored off-chain within the Farcaster network, operated by entities known as Hubs. These Hubs function similarly to Ethereum nodes, with anyone having the capability to run a Hub. They ensure a consistent view of the network by syncing and exchanging messages.
The first application built on Farcaster is “Warpcast”, a Twitter-like platform. However, the potential to extend beyond this is far-reaching, with the possibility of other social apps, such as Instagram, YouTube, or Substack, leveraging Farcaster’s decentralised protocol. This system enables users to transfer their followers between apps, preventing apps from monopolising users’ social graphs and facilitating novel experiences that merge social activities with on-chain data.
Farcaster’s client-server relationship also sets it apart. Unlike Twitter, where a single client interacts with a centralised server, Farcaster allows for multiple servers, each offering diverse features. This unbundling of client and server reduces the risk of excessive power accumulation by any single entity, echoing the flexibility seen in exporting Gmail contacts to Outlook or moving assets between cryptocurrency exchanges.
The Farcaster protocol serves as the foundation, enabling developers to build competing clients on the same protocol. This approach is reminiscent of how different services, such as Substack and Mailchimp, are built on the Simple Mail Transfer Protocol (SMTP).
As a user-centric platform, Farcaster utilises the Ethereum blockchain to create a decentralised registry for user identities. Contrary to the need to generate a new wallet address, new users register a Farcaster ID (FID) with the Farcaster contract using their existing Ethereum address. This FID is linked to their username, and while it is unique and serves as a permanent identifier within the Farcaster ecosystem, it is not an NFT in the traditional sense as it is singular and non-transferable per account. This system minimises the need for blockchain interactions for everyday actions, improving the user experience by avoiding regular gas costs.
Farcaster’s innovative approach to social networking includes three core layers: the Identity Layer, Data Layer, and Application Layer. The Identity Layer, based on Ethereum, manages operations and authorisation. The Data Layer stores authorised information, while the Application Layer consumes this data.
The protocol addresses several key challenges of decentralised social networks:
Farcaster’s implementation of these solutions has led to the development of various applications, including Alphacaster, Discove, Jam, Opencast, Warpcast, and Yup. These applications leverage Farcaster’s protocol to offer diverse, user-centric social experiences similar to traditional platforms like Twitter but with enhanced user autonomy and network portability.
Hubs form a distributed network of servers essential for storing and validating Farcaster data, crucial for both reading and writing within the Farcaster ecosystem. They initiate operations by syncing with Farcaster contracts on the blockchain to recognise every user’s account and keys. The process for a Farcaster message includes its creation, signing, uploading to a Hub, validation, and then distribution via gossip to peer Hubs.
Validation checks for a valid signature and compliance with protocol specifications. Storage conflicts, like duplicates or limit exceedances, are resolved using CRDTs for deterministic outcomes. Hubs distribute messages using the gossipsub protocol and ensure data integrity through periodic diff syncs with peers, achieving strong eventual consistency even after disconnections. However, messages may arrive out of sequence.
There is no need for Hubs to reach consensus among themselves on the content, as the on-chain components ensure the integrity and ownership of messages through cryptographic signatures. In cases where two Hubs have conflicting messages, the on-chain data can be used to verify the authenticity and origin of the messages, resolving conflicts based on the immutable records on the blockchain. Hubs also employ peer scoring to assess behaviour, maintaining network integrity by sidelining peers who fail to meet standards.
Farcaster recently introduced a major feature called Farcaster Frames. A Frame is a standard designed to enhance user interaction by seamlessly integrating external content within the app’s interface. It turns casts into interactive mini-apps within Farcaster. Frames blend static or animated visuals with interactive elements. This allows users to interact with various external links and applications without exiting the app. These interactive buttons, upon activation, dispatch a POST request carrying a signed payload, facilitating a range of dynamic user engagements. Central to Frames’ operation is the EdDSA authentication system provided by Farcaster, ensuring a seamless and secure user experience across various platforms without necessitating application switches, thereby circumventing potential mobile operating system issues related to deep link redirects.
The introduction of Frames has led to a significant increase in Farcaster’s Daily Active Users (DAU), with developers rapidly adopting the platform to integrate native crypto functionalities like NFT minting and gaming. This surge in adoption shows Frames’ potential to bridge the gap between separate, siloed experiences, offering a unified and immersive user experience. Farcaster is actively developing enhancements for Frames, focusing on features such as text input, on-chain transactions, and improved privacy measures.
Farcaster, a nascent platform in the realm of decentralised social networks, is fostering a vibrant ecosystem of diverse applications, each catering to different facets of digital interaction and online community building. Below, we outline a couple of the exciting applications that are utilising Farcaster in unique ways:
Warpcast is a flagship application in the Farcaster ecosystem, mirroring the user interface of traditional social networking sites like Twitter. This Web3 app, accessible via mobile and web browsers, allows users to share posts (casts), engage with others, flaunt their NFT collections, and seamlessly integrate their on-chain activities into their social feeds.
Paragraph reimagines publishing in the decentralised world. Drawing inspiration from Substack in the Web2 space, it enhances user experience with extensive customisation options, automated email workflows, in-depth analytics, and collaborative tools for teams. Its unique offerings include minting, NFT memberships, token-gating, and direct integration with the Farcaster social graph, unlocking new avenues for content monetisation and audience engagement. When users sign up with an Ethereum wallet, Paragraph immediately taps into their Farcaster network to recommend newsletters from their connections. It allows users to subscribe to their followers’ newsletters, view a feed of their followers’ long-form content, and access trending posts shared within the Farcaster community. Paragraph also aggregates discussions from Farcaster directly onto the relevant newsletter posts, centralising conversations and enhancing engagement. The Paragraph team recently integrated Farcaster Frames to streamline the newsletter experience, offering users a one-click subscription to newsletters, and users can read the post directly from their Farcaster feed.
Kiwi News has emerged as a crypto-focused media dApp where community participation is centred around an NFT pass. Users can share and vote on a plethora of content, such as podcasts, news articles, and videos. Kiwi News, available as a web extension and app, offers an ad-free experience devoid of social noise, emphasising community-curated content.
Wield is a wallet that offers a unique club membership model, enabling users to earn rewards through crypto transactions. It leverages the Farcaster Layer 2, Cast, for efficient message submission to a federated, open-source Hub. This approach eliminates on-chain storage fees and Farcaster ID fee requirements. Farquest and Cast are applications built on the Wield platform that enhance the Farcaster protocol experience. Cast is a prominent Farcaster client with roughly 15,000 daily active users, while Farquest gamifies the protocol, allowing users to explore it in an RPG style and earn rewards.
The Farcaster community has developed a series of utility-focused applications, including:
The graph provides a view of user registrations for the protocol. Initially characterised by a modest uptick in registrations, the graph delineates a marked escalation post-July 2023 as it marked a significant pivot to a permissionless paid subscription model. The narrative above the graph is imbued with a sense of expectancy for 2024, reflecting the bullish stance on the future trajectory of user engagement with the protocol.
These graphs offer a granular view of the dynamics of interaction within the Farcaster community, highlighting content-sharing patterns and the corresponding engagement levels across different types of casts. The first graph tracks three key metrics: links, reactions, and casts. Peaks in the graph suggest sporadic surges in activity, with notable spikes, particularly in the reactions component, which indicate viral content or events prompting increased user interaction.
The total reactions of casts after 24h graph provides a breakdown of the volume of reactions received by casts within 24 hours of posting. It’s segmented into reaction ranges, such as 0, 1-5, 6-10, all the way to over 100 reactions, showing the distribution of engagement across posts.
The data points from this dashboard paint a clear picture of a platform on the rise, demonstrating substantial and growing engagement. The total of 2.8 million casts by 306,000 users signifies a robust and active user base. The weekly and monthly active casters, at 49,000 and 53,000, respectively, further highlight this healthy recurring engagement and are indicative of a platform that users return to regularly.
The user growth rate of 11% and the high Gini Index of 90 suggest a concentrated level of activity among a smaller subset of users. This could be interpreted as a core, highly engaged community driving the platform’s interaction, a common trait in emerging platforms.
The heatmap’s representation of weekly cast activity offers valuable insights into user behaviour patterns, showcasing the times when engagement is most intense. This data could be instrumental in targeting peak activity periods to launch new initiatives or features.
The ‘Users According to Casts’ pie chart and the ‘Continued Engagement’ metric underscore the platform’s success in attracting and retaining users over significant periods. The segments showing engagement for one week, two weeks, one month, and over three months confirm a sustained interest, a crucial factor for long-term growth.
All these statistics collectively indicate that Farcaster is on the right path when it comes to building and expanding its user base.
Farcaster’s ascent is contoured by notable challenges and adoption barriers, particularly regarding user experience and the broader perception of decentralisation. This task is complicated by the existence of multiple clients, each with its own slightly tweaked UI, which is a crucial factor for achieving wider adoption. This technological chasm potentially deters mainstream users, who may value interface familiarity and content richness over the nuances of the underlying technology. A pertinent debate in this context is the public’s valuation of decentralisation and censorship resistance. While appealing to a privacy-conscious demographic, it remains unclear if the average user prioritises these aspects over convenience.
Furthermore, the discourse on whether absolute decentralisation is the end goal or if a semi-centralised model of moderation, akin to recent evolutions in platforms like Wikipedia, might be more palatable, is increasingly relevant. Farcaster, for instance, adopts a sufficiently decentralised approach to reap the benefits of decentralisation while retaining the ability to scale and combat spam. This hybrid approach offers a balanced juxtaposition of user autonomy with responsible content governance.
As Farcaster scales, maintaining the integrity of its user base and community culture presents another hurdle. The platform’s ethos of fostering a friendly and innovative digital milieu must withstand the pressures of growth, ensuring that expansion doesn’t erode these foundational values. Farcaster also faces stiff competition from not just entrenched Web2 entities but also other protocols building on SocialFi in Web3, making market penetration and user conversion a formidable task.
Overall, decentralised social media platforms, such as Farcaster, Mastodon, Minds, and Lens, offer various advantages over traditional centralised platforms, including user control over data, censorship resistance, and increased privacy. However, they face significant challenges in gaining widespread adoption, such as:
As the application layer on top of decentralised social media protocols increasingly gains traction, the influence that large applications can amass over the underlying protocol becomes an essential but often overlooked topic. Large applications can significantly steer the protocol’s development decisions, much like Gmail’s local policy changes can impact all other email providers, even within a decentralised email protocol. For a decentralised protocol to remain truly democratic and resistant to the influence of large entities, it must either have mechanisms to enforce rules against a large entity or support a competitive environment with a multitude of players that prevent any single one from forcing changes unilaterally.
As we peer into the future of decentralised social media, we stand on the brink of groundbreaking shifts poised to fundamentally transform how we interact online. These advancements herald a new era where our virtual experiences are not only more engaging and personalised but also crafted within a framework that champions user autonomy and collective governance.
Decentralised social media protocols will enable a wide array of secondary innovations. Exciting areas that we are looking into are games and commerce. Prospects include the development of games that leverage user engagement and wallet activity. These could echo the early days of Facebook gaming but without the notorious spam issues. The commerce aspect is equally promising, with opportunities to purchase NFTs and access specialised services directly through Farcaster. Importantly, these developments on Farcaster come without the risk of situations like Zynga vs. Facebook, where complementary solutions built on top of centralised platforms can suddenly turn into competitors, highlighting the platform risk associated with centralised platforms.
Although decentralised content moderation also requires a delicate balance between freedom of speech and preventing abuse, the difficulty lies in distinguishing between acceptable and unacceptable content, a task complicated by cultural differences and subjective interpretations. While no universally agreed standard of acceptability exists, democratic processes, such as those employed by Wikipedia editors or federated social networks, offer models for decentralised content moderation that are not controlled by private entities. The future of decentralised social media may involve a democratic approach to content moderation, where users decide what is acceptable through debate and deliberation, supported by technical tools and automation. This governance should be based on ethical principles, democratic control, and accountability, moving beyond simplistic censorship resistance or dictatorial administrative powers. Currently, these functions are often bundled together in a single entity, leading to issues of bias and lack of neutrality.
The intersection of social media and artificial intelligence (AI) has been an active topic of discussion primarily due to how AI algorithms utilise user data to target ads and content. Twitter and other social networks have leveraged AI/ML models to analyse vast amounts of personal data, tailoring user experience and personalising content and advertisements. This raises concerns over privacy and data ownership as these models are trained on the users’ data, yet they don’t belong to them. Decentralised social protocols introduce a fundamentally different dynamic where users own and control their social graphs. Users can “mint” posts and content, build their own “on-chain content collections”, and support content creators through microtransactions. In the context of recent developments in AI, decentralised social models offer a unique advantage for users to train their own AI models on their own data. As Fred Wilson discussed in his recent article, users’ on-chain collections can serve as personalised data sets for training these AI models. This allows for highly personalised AI experiences, as the models are trained on data directly relevant to the user, and it respects user privacy because users can decide when and how their data is used.
A more modular approach, where different aspects of the platform are managed independently, could help address these issues. This would involve separating the baseline protocol, which should be neutral, from the moderation layer, which could be tailored to the needs and preferences of different user groups.
In the broader context of decentralised social media, this modular approach could lead to more neutral “public town squares” while still allowing for effective moderation. This would make the end-user experience more valuable, meaningful, fun, and useful, as well as “appropriate” by filtering out harmful actors and posts.
Farcaster’s expansion into broader functionalities is notable. This includes implementing verification mechanisms for genuine user giveaways and allowing Farcaster accounts to be used for logging into external applications, thus enhancing ecosystem connectivity. Despite being in its infancy, Farcaster is showing immense potential and generating excitement reminiscent of the early days of platforms like Clubhouse. Its focus on building a Web3-centric, creator-focused culture is evident. As Farcaster continues to grow, the questions about the scalability of its quality user base will become increasingly important. The platform’s commitment to fostering friendly, optimistic, and innovative online spaces distinguishes it in the digital landscape. Despite facing market challenges, its strategic vision and unique approach to decentralised social networking inspire optimism and anticipation for its future developments.
Web2 platforms, having matured over time, are bolstered by strong network effects but simultaneously challenged by numerous deeply rooted design problems. In contrast, Web3 social media platforms are nascent and often face criticism for their less polished user interfaces. Yet, they embody the promise of a better and more equitable future, aiming to dismantle the ‘walled gardens’ of traditional social media. As this new era of social media unfolds, it will be shaped by the collective efforts of its users and the ethical frameworks they establish, striving to create a more transparent, accountable, and inclusive digital ecosystem for all without the redundancy of past models.
We’re incredibly excited to see the emergence of new models and primitives in this sector. If you’re in the process of creating or planning to develop something innovative in the new era of Social, please don’t hesitate to get in touch with us.
We would like to express our sincere thanks to Anthony from Seed Club Ventures, JC & Nico at Wield and Katie from Archetype for their invaluable feedback on this report.