A New Internet

Favorite Mirror Reads Archive

27 min readAug 27, 2023

On Web3 Primitives & Social Networks

Joel John

and

Siddharth

14.06.2023

Hey there,

Today’s newsletter was co-authored by Sid — who is ironically, taking a well deserved break far away from social networks. Somewhere in the hills in Himalayas. We have been brainstorming about what the state of the internet looks like & whether web3 primitives have a role to play in its future over the past few months. This piece, is by no means a market-map. It barely mentions any tokens. It is not a case for absolute decentralisation of everything on the web.

Instead, what it addresses are the incentives & reasoning behind social networks emerging the way they are today. Along the way, I break down how we may have a new vision for the internet and the creators on it through the rails blockchains enable. This is a long read that may take close to 40 minutes. You may need coffee. But I hope you take the time to read through, criticise or build upon what you read.

Drop me an e-mail if you have strong thoughts post reading this. I would love to feature them to our readers.

In 2020, when the pandemic lockdowns kicked off, I began spending an incredible amount of time on Clubhouse. I racked up a nice audience spending an hour every morning talking about what is going on in crypto. Everybody was working remotely; people were stuck at home, and much like AI today, DeFi and NFTs were the hot new thing. However, Clubhouse went from being valued at $3.4 billion to the app nobody talks about anymore in no time.

There is a myriad of reasons. The novelty faded off. People simply did not have the time to be online, plugging into conversations all the time. Or there were better conversations to be had in person. One could argue Twitter Spaces simply replicated Clubhouse for many users. But looking back on it, there’s a valuable lesson for anybody building an audience online.

You are only as good as your link to a social graph is. And your social graph is only good as long as it can be maintained and evolved. This is the difference between a city (like NYC) and an in-game community where the possibility of a social graph vanishing is very real. Physical, social graphs tend to stay stickier when compared to the ones we form online.

Surveilling Social Graphs

This challenge of being only as good as your social graph was learned by telephone operators a hundred years before social networks like Clubhouse. During the advent of the telephone, it was common for independent telephone operators in communities to tie a telephone to speakers and communicate. Much like podcasts a century later, people would speak on the telephone, broadcasting across villages in real-time. Think of it as a primitive radio.

There were campaigns against advertisements on the radio as early as 1922. The image above is from Tim Wu’s book — The Master Switch.

As larger players like AT&T and Bell took over telephone lines in the United States, these smaller telephone networks and their primitive podcasts vanished because running private, niche telephone networks was no longer sustainable. This is a chapter that we will see repeatedly as we traverse the emergence and eventual death of multiple social networks in this piece.

The emergence of a new network — like railways, telephone or internet — and new communication mediums have one thing in common. They unlock entirely new methods of collaboration. The Enlightenment era and the Arab Spring were both powered by people who found new ways to express themselves. But as we have seen, new communication forms do not establish themselves overnight.

They undergo a period of evolution and mutation before acceptable behaviour on these platforms is defined. For instance, you would not behave on Reddit as you would on LinkedIn (I’d think). And to set these ground rules and play the great social game, a social graph is needed.

But what even is a social graph? Put simply, it is the network of people and their relationships on an emergent platform. A social graph can be formed outside a digital native platform, like a university. Or it can be powered through algorithmic discovery like on Twitter. A graph can be public, where you can see who interacts with whom.

Or, in the case of dating apps, there can be closed, 1:1 graphs where the platform can charge you to display you to more individuals. But before we understand why social graphs matter, we must understand how targeted advertisements on the internet evolved to their current state.

From Shoshana Zuboff’s — Age of Surveillance Capitalism

Monetising Eyeballs

Shoshana Zuboff in The Age of Surveillance Capitalism, refers to user interactions on platforms like Google as behavioural value surplus. Historically, a firm had limited resources that it had to employ immediately to produce the goods it sold you. Or it paid ridiculous amounts in storage fees. A pencil manufacturer had to ship pencils. Ford’s car factories had to sell cars. They could not endlessly stockpile the timber or rubber for the process.

With the advent of the internet, however, this equation changed. A player like Google or Meta could keep your data for a decade until it could be monetised for their benefit. I could go to Facebook now and download all the cringe texts I may have sent my crush back in 2011 (and so could you).

Much like I often nerd out about how blockchain data can be parsed for better consumer targeting on this publication, Google’s team considered using sensors to capture user data as early as 2000. They noticed that grabbing data from wearables and in-home sensors could help create a better user profile suited to a person’s needs. At the time, little did they know that we would be carrying watches that could do ECGs around the clock or that half of the world’s mobile devices would be powered by an operating system powered by Google (Android). There was a new resource in town. User data was available in surplus. The mechanisms to monetise it were not.

In the early 2000s, most dot-com projects were what AI websites are today: an abundance of inbound traffic with little or no business model. You could license your search engine to a larger corporation or sell sponsored ads like Yahoo did. We try that here at the blog too — and as anyone would know, a bear market is the worst time to sell ads. So Google had to find a different way of selling ads altogether.

Instead of allowing people to bid on and list ads based on their assumptions about what the audience would click on, the data scientists at Google could measure and predict which ad would best suit which person. Instead of a brand’s ad manager working on assumptions, you had data scientists targeting users, allowing the brand to see a clear RoI on each click from Google.

The perfect storm was in place for the evolution of the web. A business was realising the possibility of generating and storing a resource (user data) with next-to-no marginal costs and had the pipelines (targeted ads) to monetise it. All that was missing, in the style of most venture-backed companies, was a mechanism to scale it. This is where social graphs came to play.

Quoting from the book (yet again)

At that early stage of Google’s development, the feedback loops involved in improving its Search functions produced a balance of power: Search needed people to learn from, and people needed Search to learn from. This symbiosis enabled Google’s algorithms to learn and produce ever-more relevant and comprehensive search results. More queries meant more learning; more learning produced more relevance. More relevance meant more searches and more users

See the bit about ‘more users’? One of the strongest on-ramps for more users was the network effects of new-age networks like Facebook and Twitter. Social graphs serve two purposes. Firstly, they took the internet from a weird niche technology where you had LimeWire or AOL to the cool thing kids spoke about at school. Secondly (and more importantly), it gave the Internet a business model.

More users effectively meant you had a critical mass which could be divided and sold all kinds of goods. Users belonging to similar social graphs could be bucketed and served similar content. This became the basis for the algorithmic feeds in which we currently find love, jobs, giggles, despair & hope.

As a creator, one spends time on platforms like Twitter or Facebook today as they are the means for distribution This occurs because these social networks have a never-ending appetite for content that feeds into a relevant social graph and keeps users engaged. If a community of fintech fanatics were constantly given content from crypto influencers, they would eventually be outraged and depart the platform.

Similarly, if my Web3 content were shared with an audience subset who hates it, my incentives to create content would drastically decline. Platforms play a critical role when targeting relevant audiences with content based on the data they have on users. The longer platforms can keep a user hooked, the more ads they can sell and the more data they can gather. As they hoard more data, the more relevant their ads get. This process is an infinite money-printing machine for all intents and purposes.

In Web2 networks, the social graph is the moat. If you allow users to interact with a graph via third-party applications, your chance of capturing user data diminishes. After all, the user then won’t be on a product that you control. If users can simply port their network of friends and family to a different application, they will have no incentive to return to yours, either.

We don’t have a shared protocol for social applications that work at the scale of Meta or Twitter because of how incentives are structured for existing behemoths. A Web2 product with open social graphs opens itself to competition and declining revenue. Both of which may not be a desirable outcome.

Composable Social Graphs

Centralisation of social networks in the gilded age of platform censorship also comes with risks. A recent article on The Verge sums up the question everyone has been asking amidst Elon Musk’s antics on Twitter and the United States’ desire to rein in TikTok’s growth.

An excerpt reads as follows:

But if our current social system were decentralized, you’d be able to post a picture on Instagram, and I could see it and comment on it in the Twitter app. Your friends could read your tweets in their TikTok app. I could exclusively use Tumblr, and you could read all my posts on Telegram. Different apps would have different strengths, weaknesses, moderation policies, and creator tools. Still, you’d have the same followers and follow the same accounts no matter your platform. There would be no ‘Facebook friends’ and ‘Twitter followers.’ The social graph and the product market would split completely.

What they are describing is the composability of a social graph at scale. A mechanism for users to access their network across applications in formats they deem best. What would that look like? Many applications that emerged after the early 2000s have not yet had a standardised protocol. There is SMTP for e-mail. There is DNS for resolving domain names. There is RSS for articles.

But what if you wanted to send vanishing images across Snapchat, Whatsapp and Instagram? What if you could have Twitter content with proprietary algorithms tweaked to your preferences? Or what if there were a version of Instagram that did not force you to watch reels?

Control is lost without a protocol for social graph maintenance and portability. Users can no longer determine how and what they consume. With RSS feeds, the user is in control. But with Twitter, Elon Musk and his minions are in control.

The solution to such a situation was proposed as early as 2007. OpenSocial was a collaboration between multiple large social networks to create a group of APIs that would allow platforms to replicate a user’s social graph elsewhere. For users, it meant not having to worry about adding friends again with each new network they joined. For platforms, it meant not having to compete against the network effects of an incumbent. Everyone’s winning, right?

Well, not really. As we all know, platforms have walled off social graphs today. The product was initially used by Orkut only and eventually saw over 350 million users. Legend has it that Google onboarded several social networks like Friendster and Myspace under an NDA, then broke the news to Facebook and forced Facebook to join.

For a moment, this strategy worked, as the graph below shows. By late 2007, OpenSocial’s network of applications had five times Facebook’s traffic. In 2008, there were 350 million users on the network, but by the 2010s, it became clear that an open graph was not what the internet desired. Much like Libra in 2022, a group of large organisations working as a nonprofit tend to be beaten by smaller incumbents that can move at speed.

Facebook dominated in a few years as it had cracked a critical mass of users. Which they managed to do by becoming an open platform third-party developers could deploy applications on. In the early days of social networks, people were not coming to them for content alone. Applications were a huge part of the draw towards them. Remember Farmville? The firm behind it (Zynga) grew on the network effects that came on a platform like Facebook. Each action you took in the game propagated across your social graph, which meant more friends to play with.

During the early days of social networks, applications empowered platforms to grab attention while user-generated content slowly emerged. Posting outrageous comments on the internet was a habit yet to be formed. The dopamine hits from the like, and retweet buttons had yet to be discovered. However, this trifecta — of powerful applications, network effects and distribution that came with a social graph — enabled social networks to establish themselves by 2010.

In hindsight, all the things we have been exploring in Web3 are aspects the internet has already tinkered with. Social graphs being portable? Done. Applications embedded with your identity? Yeah, I tried that. How about a single protocol that multiple applications can interact with? Boring.

There is no novelty in these new approaches, but the technical layer to enable them did not exist in the past. That change in infrastructure — from server-side, centralised ownership by monopolies to blockchain-based decentralised ownership by users is what is “new” about Web3. OpenSocial’s last update was in 2013. Nobody I know has access to their Friendster or Myspace social graph today. You cannot build or ship applications on top of Twitter like you once could. Blockchains may meaningfully change that equation.

Yield Farming Human Attention

Siddarth Jain had a beautiful metaphor to paint here. When a tree in the jungle dies, it has continuity, contributing to the growth and sustenance of other trees that come long after it. When a community on the internet dies, there is little that it passes on to what comes after it. Going back to how I started this piece, Clubhouse went from the app we all started our mornings with to one nobody cares about anymore.

As I write this, Naval’s Airchat is making the rounds on Twitter. I am excited about it because it uses AI to allow people to converse in their native languages. I would love to host our readers worldwide, speaking in their own languages as the app goes live. But when we start on Airchat, we start with a blank slate — a non-existent social graph.

Lens Protocol offers an alternative to this situation. The essence of their offering is simple. You have a social graph linked to your identity, which a wallet owns. The wallet lets you sign into a suite of apps, each serving a different purpose. In a hypothetical example, that would mean subscribers of this blog would also be able to opt-into seeing things I post on an Instagram-like feed or short-form short content like what’s on Twitter instead of separately signing up for each.

This protocol approach for human attention is new in Web3. It has worked for NFTs with SeaPort and DeFi liquidity, as we saw with Uniswap. But can human attention be shared across applications if captured in a protocol like Lens?

I don’t quite know, but there are benefits to doing so. It considerably empowers competition in social networks by reducing the entry barriers to creating new social networks. Founders could focus time on the application itself instead of bootstrapping a user base.

In such an instance, you could own your network of friends, but you would reach out to them and post content through a third-party application. Nikita Bier shared a modular approach to enabling social networks on Twitter recently. I presume he is not much of a fan of Web3, but the elements he has covered as “Reusable” are precisely the things that can go on the chain.

Image is from Nikita Bier’s twitter.

As Lyn Alden pointed out in this post, we have had open money for quite some time. But open social networks have not yet scaled substantially. Part of the reason for this is that there is a clear lack of business models. When social networks like Facebook took off in the mid-2000s, years of perfecting the advertisement-driven model had already occurred.

The means of monetising Web3 social products is not entirely clear. Now, there are a few distinctions to make here. Firstly, decentralised social networks have existed for a while with no tokens. Mastodon, Nostr, and Bluesky are all functional products without tokens. I don’t quite believe that tokens are the holy grail for the future of social networks.

Secondly, decentralisation brings challenges that might not be solvable with things as they are. Data must be stored in a decentralised social network in a P2P network like IPFS or Filecoin. That incurs costs along with it. Even if these costs are minuscule, they will discourage many users. Furthermore, no clear models exist for discovering content or algorithmically targeting users if the content is entirely on-chain.

Discovery today happens through products with huge moats on on-chain data analysis, like Nansen or Covalent. Lets ignore for a moment the fact that content is different from transactional data. They incur costs while parsing and categorising content that emerges on-chain. Who bears those costs? This ignores the fact that in such a model, a service provider can still tweak the algorithms to suit their agenda, leaving the user little choice regarding what kind of content they consume. So, we end up making the same mistakes all over again.

(I am skipping through a lot here about where user data will be stored and the privacy benefits such a model could bring users. We will speak about it in another piece.) .

What I’m trying to get at is the following:

Decentralised social networks have existed for a while. Humans are creatures of convenience. The incentives of distribution and discovery are far more efficient in Web2 native products. And there are no upfront costs to the customer. It is why much of the social graphs we know exist in walled, centralised social networks today.
Slapping tokens alone won’t compensate for the early-stage liquidity of human attention because, unlike NFTs or capital in DeFi projects, attention can’t be parked on a product. When a user parks $1,000 in Aave, the transaction may take 10 minutes. You cannot give away tokens and expect users to spend 1,000 hours on a social network. This is the reason why most social networks in Web3 die really quick deaths. (Remember Steem?)

Embedded Social DApps

So what exactly is the point of Web3 social networks? Is it just fugazi for the sake of issuing tokens and pretending as though we are at the precipice of a new internet? Or do these primitives hold promise? One way to think of it is through the lens of what @mhonkasalo mentioned in this post.

Applications require a threshold amount in liquidity to become relevant. In Uniswap’s case, it is capital locked. In the case of Mirror or Lens, it is the number of people creating content and engaging with it. At its crux, compared to Mastodon or Nostr, a token-based network can have drastically better chances of bootstrapping initial liquidity to become relevant.

From Mhonkasalo’s Substack post.

This is not to discount the possibility of airdrop hunters posting spammy content and engaging with posts for the sake of an airdrop. If you think of it, somebody like Ben Thompson (of Stratechery) or Packy (from Not Boring) has very little incentive to move to a new Web3 native platform. Their audience base is strongly embedded in their mailing lists and Twitter.

But for a new creator building an entirely new audience base, tapping into a community of airdrop hunters on Lens could be a powerful strategy. Token networks help distribute social graphs like those on Lens from 0 to 1. One instance of a creator scaling along side a platform is that of Bill Bishop. He was one of Substack’s first writers & scaled his newsletter substantially alongside the growth of the platform.

The challenge, becomes how you retain a community after you reach threshold levels — say 10,000 engaged members on an app. This is where Web3’s DApp ecosystem elements come into play. Remember how I mentioned that applications like Farmville were crucial in kick-starting large audience bases on social networks?

Applications and social networks in Web3 will have a symbiotic relationship in that neither has seen a substantial user base as they stand. But what if you could trade a token based on what an influencer you follow mentioned? Or collect an article as an NFT directly from your feed? Interfaces to enable this already exist but they are spread across applications.

Much like how Facebook empowered a generation of applications built on it to tap into their privately held social graph, Web3 DApps will be able to use emergent social graphs on protocols like Lens. The missing “bridge” here is a client layer that can enable that transition.

I am hinting at the composability of social graphs and DApps coming together. In such a case, a user could consume content and trade assets, collect NFTs, or reward creators directly without the platform taking on the risks of any of these actions. You could source liquidity from Uniswap, OpenSea, or Mirror to perform these actions.

The platform could charge a small fee (say 0.1%) for bringing together the protocol and the user on every transaction. This may seem far-fetched, but consider that Metamask alone has enabled some $3 billion in volume for swaps of assets on it. Once you have a user base, you can embed financial applications of all kinds.

This open interaction of social graphs and open-access applications is at the crux of what will empower Web3 native social networks to be a thing. As things stand today, we have isolated islands. We trade on Uniswap, often making questionable decisions alone. We track DAO activity on products like Snapshot, wondering who else is involved, and then proceed to read through Mirror and support our favourite writers by collecting their articles. Each of these is an isolated interaction in Web3. And humans don’t like sitting around alone too long.

Nobody knows which of their friends are playing which cool Web3 native game. Crypto and Web3 today is either a cut-throat, PVP game where there is only a few winners or an isolated, single-player game where you hold your assets with your dear life. The technology to enable cooperative multiplayer games are here in the form of DAOs.

But our platforms rarely ever make use of them. Think of a large crowd, putting together all the components you need to make a rocket. Then piling it behind a truck. And pushing it across town physically. All the while screaming “WAGMI” and scrolling through Twitter to see if Ethereum is a security according to the SEC. That is what we have been doing with some of these primitives.

My argument is not that Web3 native social networks will become new hotbeds for Twitter influencers to find more unknowing prey for their hot new meme coins. Genuine creators can monetise their work and empower communities to tap into it. For instance, we routinely have readers translate our work to Mandarin or Vietnamese. I love it when people take our content and make their own derivatives of it.

People often ping me and ask permission to do so to avoid drama if someone calls them out on a translated piece. One way Web3 could solve for this simple conundrum is if a person could mint a NFT (on Mirror) for the pieces we write & then upload their on NFT as a derivative of our work to the same collection. (Sidenote: For the moment, I have no plans of issuing more articles as NFTs, but I’ll soon be compiling all the translated works I see on the website).

Establishing relationships on-chain for creative work lends credibility to both the original article & the derivative, without stealing the limelight from either creators. Simple, but effective provenance. But what about the money?

I have been thinking about the commercial elements of being a creator. We have been testing a paywall for some of our archived pieces at Decentralised.co because Substack does not allow you to make content (free) subscriber-only. Naturally, we’ve had people paying for content in the past few weeks despite the paywall message that reads, ‘Please don’t pay.’ (No, seriously, don’t pay just yet.) I’ll share more on what’s planned at another point, but here’s how the math works very crudely.

For a creator on TikTok to make $60K, they would need 100 million views every month consistently for a year if their only income source was ads. A newsletter charging $20/month to hit the same figure would need ~250 subscribers. Nas pointed out that the numbers may be slightly off, but the underlying point remains.

Free content often gets excellent distribution, but the monetisation mechanisms don’t exist in a way that empowers creators that focus on smaller niches. We have seen Web3 offer an alternative through royalties in NFTs. The idea is that a creator can make an asset (like a painting) once, and every time the asset is traded, they get a portion of the royalties. I don’t think that model scales, as most artists without distribution may be unable to use it.

Communities as Networked Economies

What would occur instead is that communities formed around a creator would pool resources to support that creator. In a Web3-native social network, artists would simultaneously distribute their content (gathering eyeballs) and have a subsection of power users ‘collect’ it as we do with articles on Mirror today. These power users, in turn, could gather and coordinate much like a DAO.

When a creator releases a new product, the subscribers who have collected works from the person could be the first to access it. These feedback loops of incentivising community members who proactively contribute would enable micro-communities for creators. This would be when a creator could benefit from the economic activities of people they have brought together. Creators, would be the founding fathers of new digital cooperatives.

I believe this is the future of the creator economy for a reason. Creators have expanded to businesses to add to their revenue stream. The most commonly cited celebrity brands are Ryan Reynold’s involvement with Mint Mobile and Aviation Gin. But before that, Rihanna had Fenty Beauty, Jay Z had Rocawear, and MrBeast had his burgers. Historically, a creator’s revenue stream was only their artwork. Modern-day creators expand on their brands to capture more value for themselves.

But a creator may not be the best person to expand into a new product line. For every celebrity with a billion-dollar acquisition, countless influencers have launched a brand and failed. Even having a shot at launching a brand requires one to reach a certain scale and size.

Protocols like Lens allow any third party to query the number of likes or retweets a post has received. An application could then be built that curates only members who have received a certain amount in on-chain engagement to reach out to one another. Naturally, the challenge with such a system is that it would incentivise individuals to spam for engagement. But with strong moderation, such a curated social graph could be compelling if applications are built on top of it.

I try to explain what the transition would look like in the image below. With due apologies to the readers from mobile devices — the model below shows how a web2.0 influencer would differ from a community curator in Web3.0. Blockchain-enabled payment rails would enable creators to enable member-to-member commercial interactions. The green lines on the left side indicate payments between members, and the blue dotted ones towards the creator indicate potential royalty payments.

For instance, someone could build a version of the Producthunt and bootstrap community members from what we have at Decentralised.co. A third party could build an Angel List or a syndicate DAO — and query our community for the most engaged VCs and founders. Both of which are a possibility today.

This composability of social graphs is missing on today’s internet. When we run ads, we pay Google or Meta (or this blog’s authors) to mention a venture to a smaller audience subset. However, the human mind works because we have effectively blocked out advertisements from our periphery. The average person sees about 4,000 to 6,000 ads on a given day. We consume without paying active attention, and human attention has evolved to ignore advertisements because it is a cognitive load we did not ask for.

Composable social graphs can fix this by allowing people to buy a new product. For instance, if a new game is launching and they wish to tap into the Decentralised.co community, all that is needed is to list them on the Substack. Users can then choose if they wish to interact with their product. This switch — from the platform determining what is best for a user to one where users can select products based on their preferences — is the fundamental promise of what a Web3 social network can offer.

You can always argue that this seems far-fetched and unnecessary, but experimentation is the crux of what made DeFi and NFT so powerful. When centralised product managers run a platform like Instagram or Twitter, you have no say in how the product evolves. You could also argue users should not have a say in how a product evolves — but I think it is different when it comes to social networks. When users are the ones driving the growth of a platform, there needs to be a balance of power between shareholders and stakeholders.

Community-driven content networks have existed as long as the internet has. Wikipedia is a powerful example. What Web3 brings to that equation is the probability of financialisation and user ownership. Would the contributors of Wikipedia like to have a say in how the product evolves? I would think so.

Reaching large numbers of users (scale) has long been the primary incentive on the internet. As I wrote earlier, people write on Twitter instead of Mirror because the distribution is on the former. However, if we change the incentives to ones where people are no longer the product, we can form the basis for a better internet — one that does not involve creating content to spur emotions.

It may seem far-fetched to think of a social network involving payments, but Twitter already charges $10 for premium subscribers, and the internet has ample instances of a community going from free consumers to paid ones.

In India, most of my generation used to torrent content in the late 2000s because products like Netflix or Spotify were not around; even if they were, these platforms would not accept our debit cards. However, a shift has occurred over the past decade. As an increasing number of Indians came online, and the payments network within the country evolved, we had what could be considered an economy of scale. Paying for access to watch the latest movie or cricket matches became commonplace as paying the amount was easier than bothering with the illegal route. Convenience is the ultimate sales pitch if your consumer does not have to break the bank to make a decision.

Capitalising on content on the internet has been restricted to an elite few that have reached scale. Web3 native social networks allow creators to change the equation by offering new alternatives to monetise their social graphs.

Looking through this lens, we will soon have digital native nation-states. Balaji Srinivasan’s work looks at the other end of this equation — a time when a digital commune can perform functions that a conventional state does. I argue that creators will be founders of micro-nations oriented towards niches before that transition occurs.

They will not collect taxes or issue identity verification documents like the government today, but they will be crucial in establishing and growing entirely new industries. This may seem far-fetched, but consider that Satoshi and Vitalik Buterin are the founding fathers of their digital economies. Their ownership of Bitcoin and ETH represents the value they generated in creating new financial paradigms.

Power to The User

Erik Hoel is one of my favourite writers on Substack. In a recent post, he argued that a new social network’s emergence is unlikely and not worth pursuing anymore. As we scale, we reach what he calls a ‘Semantic Nadir’ — a tendency to take things in the worst possible way. Did you post something about how much you like burgers? Someone on the internet will see this as a call to war against vegans.

He believes that as human networks scale, our tendency to gossip or lash out at one another grows. The internet can curate the worst of what humans are capable of and present it to you overnight.

He is right in his argument so long as we presume that distribution (alone) is the key incentive for social networks. My argument is that incentives can be restructured altogether. However, before this occurs, there will be a period of transition. This will be when users can tweak the algorithms to suit their preferences.

In such a system, the social graph would not be user-owned, but the algorithms that determine what is shown to the user can be tweaked by the user. This may seem far-fetched, but platforms like JoinColumn* are already working towards this vision.

One place where the internet is already seeing the power of communities and users is Reddit. APIs that power external mobile apps on Reddit saw a substantial surge in pricing. The change will affect everyone from behemoths like OpenAI, which may be using data from the platform, to smaller mobile apps.

Data from Reddark. Numbers indicate size of the subreddit that has gone private in protest. 8400 of 8800 Subreddits are currently private in what is one of the largest protests in the digital realm.

The change in pricing made it impossible for interfaces to Reddit, such as Apollo, to keep functioning. Multiple large subreddits catering to tens of millions of users have begun going ‘dark’, which involves setting these pages to private so that users can no longer access the subreddit.

The protest may be somewhat weak unless users leave Reddit in droves. (At the time of writing, some 8400 subreddits off a total of 8800 has gone private). But as platforms like Snapchat and companies like Meta have shown in the last decade, social networks have strong Lindy effects. The longer they have been around, the greater the odds that they will continue to exist. This outcome is because a user faces a high opportunity cost to delete their Facebook or Twitter accounts entirely.

They cannot access the same group of friends elsewhere without much friction. Portable social graphs — like the ones enabled by Lens — offer an alternative where a user can switch out of the platform but still hold on to their network of friends.

Consider it as if social graphs were nation-states and platforms were businesses. It is incredibly hard to switch out of a nation-state entirely, as anyone who has had to move to live elsewhere would know. But a platform with commercial interests should very much be treated as an entity that can be switched at will.

The internet does not give this option to users today. We see aspects of it with text-based applications like Signal and WhatsApp. You can choose to quit WhatsApp entirely and text the same friends on Signal, then realise that only a small fraction of your friend group even uses Signal.

Ultimately, creating a new internet with entirely new incentive structures requires rethinking how the past three decades of the internet have evolved. Presuming slapping token incentives or shiny new buttons would attract users is a bad heuristic. We need creators and their audience bases to rethink how and why we interact with one another on the web and to what extent bits of it can be monetised in ways that don’t involve tapping into user data.

A model that protects privacy without distribution may not work. Similarly, one that reaches scale without retention will not work. We will have multiple iterations and narratives in the market as these transitions occur, but it is clear to me that there has never been a better time to attempt to create a truly Web3 native social network for the masses. Humans are creatures of habit. Changing habits we picked up and perfected over 30 years of consuming free content and disbursing pointless ads will be a slow, uphill batter.

One of the controversial stances Peter Thiel was infamous for in the early 2000s was his thoughts on how we were in an age of stagnation regarding tech. I’d think the same if I were a VC in that era. (FWIW, I think the same now, because mentally I am a cynical old man) In fact, Tascha from Twitter recently had a very similar stance — that crypto has not had anything novel or groundbreaking for quite a while, and until we deliver on it, the markets may not recover.

I echo those sentiments. But I also believe we are looking at the problem incorrectly. Crypto does not lack fund apps or trading products. It does not even have a UX challenge if you consider account abstraction. It lacks a social graph that can help propagate these products in a way that keeps the consumer entertained and engaged. And that will not happen until we produce social products that provide more than shilling tokens. We see large-scale, privately held graphs emerge. Layer3, for instance has over half a million users with credible, on-chain activity in their userbase. They are well-positioned to expand into a social network.

As the user above pointed out on Twitter, the difference between AI and crypto is in how many people use the underlying technology. One way to inverse the relationship between token holders and product users is by looking at social products where a token is not required to interact with the product.

Much like it took well until the mid-2000s for large-scale, retail-oriented social networks to emerge in Web2, it may take a while to see social networks of scale in Web3. It is a function of time. We have tried, experimented and failed multiple times in the past with social products in the sector. But the difference is that in 2023 the technology to enable such a social product exists. And that gives me hope.