Social Web 3
August 11, 2022
Beyond Buzzwords: Next Generation Of The Social Internet
Social media has become a major part of most people’s lives. Current estimates put usage at over 2 hours per daily on average. While growth in time spent has seemingly plateaued in recent years – perhaps in the face of increasingly negative externalities – social media platforms obviously play an important role in our day to day lives.
Figure 1. Daily Social Media Usage by Internet Users Worldwide (Statista)
Social media’s earliest roots can be traced to 1997 and a website called SixDegrees.com. This allowed users to create profiles and connections and send messages. From here complexity and sophistication increased. Hot or Not, Friedster, Myspace came in the following years, and the social web as we know it today originated in the mid-2000s with Facebook and Twitter.
The monetization models of these platforms are based on advertisement. Use the data to serve a higher likelihood of click through ads – use additional user data to better optimize these ads – repeat. The cycle is reflexive, the more they understand you the better they can serve you ads.
The toughest part of this is both the degree to which these platforms existing in the Web 2 environment are willing to go to acquire this data, and the misaligned incentives of their competition.
Each platform is inherently incentivized to capture and not share their data on you with their peers. If they can serve you higher likelihood-of-click-through ads than their peers, they can capture a greater share of two fixed resources (your disposable income and time). In their competing pursuit to do this, we have seen all kinds of questionable behaviour at these platforms.
Social media is blamed for many things including increased societal strife. What makes us wonder is whether 15th century Europe did not feel the same. Can you blame Martin Luther for getting ideas that led to a century of religious warfare? Could they perhaps regulate the printing press back then so that people do not get any ideas? Newsflash: they actually did censor. How did it work out for them
Today many advocate for moderation of social media, which is a more politically correct way to say censorship, we do not see an interference of third-party authority as a solution. So in the balancing act between anarchy and authoritarianism; we seek the solution in the arms of technological innovation. Because fundamentally “the medium is the message”.
The internet changed the topology of human communication, with existing social platforms as a mere proof-of-concepts. We’re ready to embrace even more radical technologies that allow bottom-up online tribe formation rather than top down centralized algorithm control. Because if the algo decides what is the “official truth” everyone will just fight to control the algo.
One thing about current social media platforms is true; they control the narratives and the user becomes powerless; almost a serf in the digital domain. As Chris Dixon puts it, given enough success, a given platform’s relationship with the network participants switches from positive sum to zero sum, attraction turns into extraction. Also, the cost for user to leave the platform are much higher than the platform using this one user – a perilous asymmetry.
Intuitively, we understand the suboptimal nature of the current status quo but we face a problem of coordination. Even if you choose to join a new and better platform, you will end up there alone. It only makes sense to switch when others are switching.
In the last 5 years facebook has nearly tripled free cash flow from 14b to 40b. The value creation by the platforms are often on the back of being both highly useful and providing certain societal functions we will likely not turn our backs on.
As we had mentioned earlier, the question now is not should we have these platforms, but how can we improve these experiences on both a personal and social level. While the former is doable the latter is rather hard to address as technology does not regulate human vices.
We believe that the next generation of social media will leverage technologies that are these days referred to as Web 3. Social will not stop evolving. However, to rely on purely idealism as a motivator of adoption is naive. We need a user experience that is much better than today and we need to invent tools that users will absolutely never attain in the Web 2 realm.
Web 3 Social – What is it?
We see a lot of talk about Web 3 social these days, but what does it actually mean? What do we think this looks like, and how will it work?
Breaking it down into the subcomponents we have Web 3 and social. Web 3 is a buzzword that we have chosen to describe a new generation of applications built on or related to blockchains. Web 3 is characterized by being mainly decentralized and open source. But to this day, most of the applications claiming to be Web 3 lack this attribute.
We still think that many of these applications will be “Web 3”, however, it would be more useful to think about Web 3 as an emergent phenomenon rather than a specific application. Web 3 will be a choice, or a preference happening on a very fluid technological stack.
Applications can facilitate a Web 3 experience but “Web 2 to Web 3” is mostly a spectrum that is related to a user preference. This probably depends on how much of the state and data is kept in a decentralized versus a centralized way.
In its simplest form, we can expect Web 3 social to look more like an opt-in model. The biggest feature change is owning your social graph and your data and attractive monetization models for users.
From here users can choose to opt-in (share their data) with a given platform and to what degree. We are already seeing examples of such a model with Lens and the platforms building on top. Brave Browser has been an early pioneer of such a strategy, and can be likened to netscape in the Web 3 timeline.
Perhaps even more interesting is an ability to tune social media feed algorithms. We touched on this idea briefly in the middleware thesis, but here we flesh it out more deeply.
While current models optimize for engagement (time on the platform, click-throughs, etc.), future Web 3 platforms will be able to offer flexibility of curation. In fact, given the accessibility of the data in Web 3, we view the curation of the content as the key point of value accrual. The winners here will be those providing the most desirable experience to users (i.e. a more optimal equilibrium).
This can be sliced from multiple angles such as the desired balance of personal social content (your friend got engaged) and non-social content (like news), profitability (users opting to see more ads), learning (educational content), and emotional tone (feed for happiness – more shitposting yay).
The post web 3 social world
How Do We Get to Web 3 Social
Mapping The Web & New Bundles
By introducing the PageRank algorithm Google was most successful in structuring the wasteland of unfiltered data that was Web 1. PageRank reads flows of outbound and inbound links, thereby creating a graph of websites. There were other efforts before this but none emerged so dominant.
Each website was then ranked based on its importance in the graph determined by how much traffic passed through them and how densely they were linked to other important nodes in the network.
The introduction of PageRank created the context for unfiltered data and sent the internet on a reflexive flywheel. It created momentum – popularity led to prominence which increased the popularity of websites. From this reflexive momentum emerged highly ranked websites that now serve as signposts – markers by which we navigate the internet.
If we agree that blockchain protocols are the foundation of Web 3 then I propose that the wallet addresses are the new websites that should be indexed and not the apps.
Web 2 has been top-down oriented. Websites (applications) are anchoring information – it’s an app-centric flow. Apps own users.
Web 3 is address centric. The users are embedded in the protocol and they do not require apps or websites to create accounts or online identities.
Blockchains disintermediate apps by integrating users directly into the protocol. Users own addresses, which are in fact accounts without any identifiers. Users no longer rely on services to give their identity context, instead, services rely completely on users to bring their identities with them, creating context organically and bottom-up.
Blockchains are state-keeping machines. They decouple context (state) away from the service. This is how we are creating a new Web bundle.
The addresses and underlying blockchains providing state transitions are essential to these new bundles. The app, although still providing UX access, is a different class of user. In the same way that both a restaurant and its patrons use the infrastructure of a city, both apps and users leverage the state of a blockchain.
As these addresses engage by sending transactions and signing messages, they create a complex interactive graph. The current Web 3 graph is mostly financial, tracking the flow of value between addresses.
How do we make sense of this graph? And how do we make it “social”?
Horizontal And Vertical Relevance
Web 3 as we know it today is mostly context independent: money value does not depend on anything – it is horizontally relevant. Financial capital does not require context to be valuable.
In contrast our social interaction mediated by social media is context dependent: the likes do not mean much when abstracted away from your social network. Social capital is therefore vertically relevant, not universal and definitely not transferable.
The context mostly refers to a cluster within a network. If you hold certain status within one community it is not directly transferable to a different community. E.g. you can make people ape coins, but you probably not gonna get them to buy a certain type of energy drink.
As an influencer, whether IRL or on Twitter, you are trying to transfer your vertical relevance into a horizontal relevance. This is called “monetizing your network”.
Vertical relevance metrics in the digital space come in the form of engagements such as likes, response, subscribers, upvotes etc. These are the reputation scores for our online identities. To this day, these social metrics barely live in Web 3 and are mostly bound to Web 2.
Today we see some instances of social metrics derived from the user interactions, notably DegenScore. But financial transactions cannot be the only one type of interaction weaving the fabric of Web 3. We need to add different contexts than just financial, attaching social identifiers to the Internet of Value.
The pre-PageRank internet did not have an explicit graph of apps (websites) that anchored Web 2. Web 3 currently suffers from the same problem. Separating state is good in terms of digital sovereignty but suboptimal in terms of creating a context.
What do we mean by context? – Context on the web is weaved by indexing graphs. Graph is a mathematical structure that carries information. We create this structure by attaching identifiers (values) to nodes in this structure.
Given that relationships of users in Web 2 are hidden behind the walls of applications we are not allowed to generate context independently. Everything is served to us by algorithms that leverage closed mathematical structures.
Your Web 3 address is anonymous. Its default state has no identity. This is good and bad. Good because you can remain anonymous. Bad because, by not carrying any identifiers, it cannot become social meaning it will not connect with other humans at scale.
A Web 2 account’s default identity is derived from an application. To attain an identity for a Web 3 account, we require identifier-as-a-service (identifiers is a word we use for verifiable attestations). In other words, we need apps that provide identifiers to our accounts.
As mentioned above, there is no inherent social graph for Web 3, but there is a financial one, since we can establish financial flows between accounts. Unlike financial capital, social capital is not transferable. At least not directly. Vitalik calls this property soulbound.
One way to tell one’s social capital is figuring out how many friends one has. In the world of bits – how many followers one has or at which institutions one is employed. Here, centralized institutions lend an attestation to an individual. Institutions are reputation third parties. This attestation identifies the person in order to help mediate interaction with other people.
Generally speaking, identity is predicated on non-transferability, it’s not disposable but it is fluid. If you create a LinkedIn profile, you are flexing your job history. You can delete it or change it but selling it does not make sense.
The social media of Web 3 will likely be emergent. This means it will not be built on a social media platform rather enabled by social identifiers. Non-transferrable attestations will serve as identifiers. Applications of Web 2 or 3 will become these reputation third parties.
Web 3 will be very fluid, modular and less static. It will not be app-centric but rather identity-centric. You can take your identifier anywhere. This generates downstream innovation with apps built to cater to users seeking to add identifiers which enable the next generation of social web.
The great thing about this is that you can port Web 2 identifiers into Web 3. This way, Web 2 apps can continue to provide service to Web 3, giving away some of its power. This also suggests that there is no place where Web 2 ends and Web 3 begins.
Web 2 apps can provide service to Web 3 accounts as a source of attestation. They will monetize their local state. The general state-keeping is built into the accounts that live on the blockchains or into new modular data layers like Ceramic or Kwil as the legacy platforms transition to more natively Web 3 monetization models.
Web 3 (or “Web3ness”) is mostly describing a power shift from Web 2 apps but it is not a destination. Who controls the state controls the coordination tool on the web. E.g Twitter’s role could change but it can remain an app that we will use. Your Twitter account can become a component to a richer identity of Web 3 spanning various venues.
Blue Checkmarks: Finding Relevance
We’re fond of a whitepaper published in 2017 by a now-failed project called Userfeeds who were too early. Userfeeds wanted to build a protocol that “creates an open layer for integrated relevance algorithms that combine signals from consensus layers (blockchains) but add a component of personalization”.
What do they mean by this? Let’s break it down. Userfeeds defined relevance in a mathematical sense as “predictability”:
The more relevant a signal is, the more predictive power it has about another signal. The process of identifying relevant information in X about Y is done by “compressing” information in X without losing the ability to predict Y. The end result of such “compression” is the ability to predict Y with minimum information about X…Relevant information in behaviours (x) of people on the Web (X) is the information that it provides about other behaviours (y) of other people on the Web. (Y)
When people talk about decentralized social graphs they forget that we do not really need new nodes, only need to describe the edges; meaning we only need to define relationships between users. This could be done by verifiable attestations or as we call them in this research: identifiers.
There will be no single provider of a permissionless blue checkmark. There will be various identifier providers, some will get more traction than others. Some may be more prominent among certain clusters.
Maybe unlikely but Twitter itself could pivot its business model towards a more open web dynamics by providing it attestations. Imagine “Vitalik is following me” NFT badge. They could monetize it by keeping identifiers up to date as the value of such proof would decline in time.
By porting Web 2 identifiers is one way to describe the relationship of wallet addresses. At the same time we are generating some social proofs on Web 3 already: owning NFTs, owning fungible tokens, being part of DAOs, participating in governance and so on.
All of these can be translated into identifiers without doxxing. Pseudonymity is the future of Web 3 as _“pseudonymity allows freedom after speech and defends against social supply chain disruptions” (Balaji Srinivasan: The Pseudonymous Economy)._
Today, Web 3 is a reference to a pseudonymous economy in the cloud. In the near future it will be enriched by user friendly identity management that preserves privacy of users and allows them to aggregate reputation and manage their online identity and “own their social graph”.
Diagram by _Shrey _
One of the projects working on this is Sismo – a UX friendly Web 3 identity management system based on cryptographic attestations of metadata based on zero-knowledge proofs. The user can manage aggregate of metadata from its ETH addresses and Web 2 social profiles via Sismo and selectively reveal the information through NFT badges that serve as playful attestations.
E.g. wallet X can prove that it’s followed by Vitalik on Twitter without revealing who it is.
Sismo allows you to aggregate your reputation signals into one and then use specific online personas. One might want to have a persona for governance discussions and then a different persona for other purposes.
We can think of Sismo as an identity middleware. It’s extensible and gamified as one can come up with an own idea for an attestation “followed by X + Y” or “owns X amount of token Z”.
We envision many attestation providers just like Sismo competing for the most reliable and popular identifiers. Some will be adopted, some not, others will become fashionable just for a period of time (just like Instagram or TikTok filters). Think of your selection of identifiers as your own personal filter that you are imposing on the social graph available to you.
Further to this, we have also seen something of a similar vein win the ETH New York Hackathon: ETHdos. Effectively a ZK way to private show degrees of separation across Web 3, it’s an interesting concept for building new methods of exploring connections.
Once we adopt the attestation services we need aggregators and algorithm providers. Imagine subscribing to different providers of pagerank-like indexing, and personalizing the web experience to your needs. You can switch the algo based on your preference.
This is how relevance will be defined in the next generation of Web user experience. Going back to our point of the importance of curation, because of the distributed nature of these attestations, you need the interpretation layer to make sense of this content and deliver a tangible experience
Another important aspect in improving coordination is to build applications that will help in forming common knowledge:
_“Common knowledge is often important for coordination. For example, a group of people might want to speak out about an issue, but only feel comfortable doing so if there’s enough of them speaking out at the same time that they have safety in numbers.” (Vitalik) _
If we can really achieve this we not only build a private and permissionless social graph but also begin to form an open-source knowledge graph, a dynamic user-centric catalog of ever-changing information just like Wikipedia but more fluid and potentially cryptographically verifiable.
To recap, we have talked about previous Web 2 social media models, what Web 3 social will look like, and how we get there from the current state of Web 3. We see a world of attestations/soulbounds/nfts adding the vertical context necessary to make Web 3 socially relevant
To better illustrate how Web 3 social will differ we provide a simple illustration:
Web 3 dating today:
Web 3 dating of tomorrow:
Social internet of Web 2 was built on network effects and slowly transitioned into a newsfeed-like content serving. The future is long-tail social media. The platform will not remain fundamental but could transform itself into an optional reputation ledger via verifiable attestations aka identifiers. The future of news feed and content is cross-platform aggregation.
One important thing to note is that we cannot expect that Web 3 social is a panacea to the ills of social platforms. This is probably rooted in something more fundamental. It is always good to keep in mind that “the medium is the message”.
Disclosure: Zee Prime has invested in Sismo.