Future of Social Networks (1 of 3)

In 2017, a group of MIT Media Lab researchers claimed in Wired that decentralized social networks “will never work” [1]. In their piece, they cited three impossible challenges: (1) the question of onboarding (and retaining) users from scratch, (2) the (mis)handling of personal information of users, and (3) lucrative user-targeted user advertisements. In all three cases, they argued, incumbent tech giants, such as Facebook, Twitter, and Google, simply had too far-reaching economies of scale to make room for any significant competition.

Fast forward half a decade later, what was once hailed as “impossible” seems no longer so far-fetched, and we seem to be on the dawn of a paradigm shift in the way we conceptualize social media networks. In this three-part series, we will examine how new ideas in decentralized social (DeSo) seem to address these “age-old” questions, specifically, (1) the use of open social graphs in solving the cold start problem, (2) using proof-of-personhood and cryptographic techniques to solve the userhood problem, and (3) leveraging tokenomics models and incentive structures to solve the revenue problem.

Social Graphs and the Cold Start Problem

Social media platforms invariably face the cold-start problem: attracting and engaging users from scratch without an existing user base or network effects. Traditionally, nascent social media startups such as Snapchat, Clubhouse, or recently Threads have sought to overcome this through brute force and sheer marketing prowess. By capturing everyone’s attention at just the right moment, whether that is through a novel UX design, media headlines, or FOMO, they launch a huge blitz of signups to almost instantaneously build up a moat of users on the platform. For example, in a matter of 5 days, Threads was able to onboard a mind-whopping 100 million users [2].

But more often than not, these successful marketing campaigns are met with an existential crisis: how do you retain all of these users, and continuously generate new content (and profit)? This is the problem that Clubhouse previously faced, and Threads is currently facing. And as these applications die out, the lucrative user social graphs and profiles that these platforms build die out with it, so that future aspiring social media networks need to repeat the difficult marketing stunt all over again to bootstrap their network.

An example of a social graph [3]

The fundamental problem behind all of this is that in web2 social networks, the social graph layer (which annotates the relationships between users) is inseparably wound-up with the social application itself, such as Facebook, Twitter, or Instagram. The two layers are symbiotic: the novelty of the application bootstraps the social graph, which in turn acts as the primary moat of the social media application. Despite all their problems, the reason why users don’t leave Facebook, Twitter, Instagram is simple: all our friends are on it.

But what if we decouple the social graph and the social application? What if, even after Clubhouse (or Threads) dies out, we can still make use of the social graph created in their glory-days to easily bootstrap another social application? This is web3’s response to the cold start problem.

Using Public Blockchains as an Open Social Graph

vitalik.eth on Etherscan

In a certain sense, a public blockchain like Ethereum is itself a social graph. If I look up an ENS domain or a person’s wallet address on Etherscan, I’m able to inspect that person’s on-chain social profile: what assets they are holding, who they are transacting with, and infer what communities they are a part of.

This on-chain social profile seems to be a natural jumping off point for a new decentralized social network, and indeed, is a path which several companies seem to be exploring.

a16z profile on Debank. Data as of July 29, 2023.

Debank, for example, transforms the hexadecimal dump on Etherscan into a human-readable portfolio or “profile,” and offers the functionality to send direct messages to these different portfolios, thus using this on-chain data as a way to bootstrap a direct-message style social network. A similar route is taken by 0xPPL, which seeks to also use on-chain user profiles to build a Twitter-style social network. This general strategy of making raw transaction data readable and interpretable to “layman” users has been accelerated with the maturity of sophisticated Large Language Models such as GPT-4. Cymbal, for example, purportedly uses GPT to generate conversational summaries of transactions and trends to create a hybrid between a data-resource, news feed, and future social network [4].

Building Native Social Graph Protocols

A problem with simply relying on public blockchain data, such as on Ethereum, is that the data is simply not rich enough for social applications. Because public blockchains were built first and foremost with financial applications in mind, rather than social applications, most of the natively collected data on-chain, such as transaction history, account balances, and token data, are not necessarily the most useful to a social network.

Overview of Lens [5]

Instead of simply using the native on-chain data as a social graph, one idea is to build a new, dedicated social graph protocol on top of a public blockchain. Lens Protocol, for example, makes use of the key observation that across social applications, there are common denominators for social interactions, which they then abstract into different on-chain actions such as “posting,” “commenting,” and “mirroring” (i.e. sharing or re-posting) [5].

Farcaster has similar abstractions on its social graphs, such as a “cast” (post), “reactions” (likes), and an “amp” function in which users recommend other users that they believe are worth paying attention to [6]. The primary difference between Farcaster and Lens is on their technical implementation – whereas Lens places everything on the Polygon blockchain, Farcaster places its ID registry on Ethereum itself, and runs its social graph on an L2 as a delta graph.

A third notable social graph protocol is Cyberconnect, which has a greater emphasis on link aggregation (of both on-chain and off-chain sources) through its link3 mechanism, as well as focusing on Events and FanClubs as initial use cases.

Crucially, for these social graph protocols, they are not necessarily building the top-layer social applications, such as Twitter, Facebook, or Instagram. Rather, they provide the open social graph layer (essentially an SDK) necessary to quickly build and scale these top-layer applications. As mentioned previously, the core advantage of this is that even if a once-successful social app dies out (Clubhouse-style), the social graph generated can still be used by other developers. Thus, only one marketing blitz, or one successful app is necessary to bootstrap the entire ecosystem.

Designing Decentralized Social Media from the Ground Up

A third strategy in onboarding is to build a decentralized solution from the ground up. The premise for this is that social media applications are such a cornerstone of our digital experience that there needs to be a dedicated blockchain (or other decentralized) solution that puts the primitive actions of a social media application natively, rather than channel it through a protocol built on top of an infrastructure originally designed to support finance use-cases. In short, we need some sort of a social-media “appchain.”

DeSo Home Page

One of the most notable projects following this strategy is the DeSo, which is building a L1 blockchain dedicated to social applications. Instead of focusing on “transactions per second” like other mainstream public blockchains, DeSo seeks to optimize for “posts per second,” as well as social applications’ need to handle both communication and storage, which general-purpose blockchains like Ethereum are not necessarily optimizing for (think about all the images and videos stored on something like Twitter and Instagram). On top of this L1 blockchain, DeSo plans to build a wide selection of social applications, including long-form content (like Substack), short-form content (like Twitter), and Reddit-like applications [7].

Other decentralized social media platforms, like Bluesky and Mastodon, can also roughly be following this strategy of designing decentralized social media from the ground up. Strictly speaking, they are not blockchain based solutions, and instead rely on a federated server system to ensure that posts are sufficiently decentralized. Mastodon, for example, uses a system similar to email, where users can choose between different service providers (like Gmail, Hotmail, or iCloud). In the same way that an organization is able to set up and customize its own mail server, each “instance” on Mastodon will be a self-regulated and customizable community [8]. Bluesky, on the other hand, is an application developed on the open source AT Protocol, which is essentially an open social graph with APIs such as “follow,” “like,” and “post” that is optimized for a Twitter-style social-media platform [9].

The commonality between DeSo and projects such as Mastodon and Bluesky is that they reject the notion that the existing public blockchain designs (epitomized by the EVM) are suitable for a social network. Although this approach undoubtedly gives these projects more fine-grained control design decisions and user experience, in doing so this strategy severs the potential connections and cross-pollination with DeFi, existing NFT communities, and other mature elements of the web3 ecosystem. Moreover, it remains to be seen just how “decentralized” these solutions are, especially in an environment where their decentralization is not guaranteed by a public blockchain. Will these solutions, in the end, once again bundle up the social graph with the social application, just like in existing social networks, or will they sufficiently decentralize the social graph layer and attract a wide range of applications and developer teams? This is a key question for the future of web3 social.

References

[1] https://www.wired.com/story/decentralized-social-networks-sound-great-too-bad-theyll-never-work/

[2] https://www.theverge.com/2023/7/10/23787453/meta-instagram-threads-100-million-users-milestone

[3] https://www.businessinsider.com/explainer-what-exactly-is-the-social-graph-2012-3

[4] https://decrypt.co/149202/cymbal-human-readable-ethereum-blockchain-explorer-etherscan

[5] https://docs.lens.xyz/docs/overview

[6] https://hackmd.io/IP-8snyMQfOGxV3LUjlJbA

[7] https://docs.deso.org/deso-roadmap

[8] https://techcrunch.com/2023/07/24/what-is-mastodon/

[9] https://atproto.com/guides/applications

Disclaimer：

1. This article is reprinted from [veradiverdict], All copyrights belong to the original author [PAUL VERADITTAKIT]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.