# Enshittification Still Matters

> DeFi Is Not Enough

**Published by:** [DFern Longform](https://paragraph.com/@dfern.eth/)
**Published on:** 2026-07-21
**URL:** https://paragraph.com/@dfern.eth/enshittification-still-matters

## Content

In January, Ryan Sean Adams (@RyanSAdams) made a fatalistic argument (as a subtweet to Merkle Manufactory passing the torch on Farcaster): after ten years, Ethereum's use case is finance. Assets and money verbs. Stop getting distracted by sidequests. Focus on DeFi. DeFi will change the world. DeFi is enough. There is a great deal to agree with here. DeFi is important. Stablecoins are important. Payments, tokenization, prediction markets, and open financial infrastructure are all real advances over the closed systems they are beginning to replace. But DeFi is not enough. Not because every crypto project needs to promise a science-fiction future. Not because we need flying cars, digital land, and a full metaverse before we can call Ethereum successful. The case for decentralized social is more practical than that. The internet runs on a handful of platforms that own the identities, audiences, distribution, and economic relationships of the people who use them. That arrangement works tolerably well until it does not. Accounts are hacked or arbitrarily suspended. Algorithms change. Monetization is turned on, then off. A creator spends a decade building an audience and discovers that the audience belongs to the platform after all. This is not a sidequest. It is part of the core problem crypto was supposed to address: how to give users a durable claim on things that matter to them online. The Wrong Lesson From Base Jesse Pollak's (@jessepollak) recent retrospective on Base was candid in a way that crypto leaders rarely are. He owned the fact that Base's push into creator coins and onchain social did not produce the adoption he expected. He described the social side of the market - Farcaster, Zora, mini apps, and creator coins - as having “disintegrated completely,” and concluded that better money is enough to bring a billion people onchain. The accountability is welcome. The conclusion is not. Social did not fail. A particular product thesis failed: that an outsized push into content and creator coins, often treated as the default answer for every creator and application, could be the center of gravity for a social ecosystem. That is a much narrower claim than “decentralized social does not matter.” The distinction matters because content coins were never synonymous with social. They were an experiment in financializing attention. Some people found value in that experiment; many did not. The problem was not that creators, communities, and communication have no place onchain. The problem was mistaking one speculative mechanism for the entire social layer. The hosts (@age_shulman, @nounishprof) of GM Farcaster (@gmfarcaster) made the obvious point during their discussion of Pollak's post: crypto is inherently social. Bitcoin was introduced to a mailing list. Every token, protocol, market, and application is discovered through people talking to one another. Prediction markets are social. Perpetuals are social. Memecoins are almost entirely social. Even payments are frequently driven by social coordination and trust. It is particularly strange to declare social a failed bet in a long post published to X. The message reached millions of people precisely because social platforms are where crypto narratives form, capital coordinates, builders find users, and communities decide what is worth paying attention to. Better money is necessary. But money does not distribute itself, explain itself, or build relationships around itself. Social Is Infrastructure The strongest case for decentralized social is not that it will immediately replace X, TikTok, YouTube, or Instagram. Network effects are real. A new social network with a technically superior protocol but no friends, no conversation, and no cultural energy is still an empty room. The case is that a few social-media companies should not be the sole owners of the identities, audiences, and relationships that users create. Those companies can serve users and advertisers, but users should still be able to take their identity and social graph elsewhere when a platform changes the rules. On the traditional web, your identity is a revocable account. Your follower graph is a database entry. Your posts are available only under terms that can change whenever the owner chooses. The platform can make you visible, invisible, or unavailable. It can extract more value from creators and users as it becomes harder to leave. Nikita Bier, X's head of product, illustrated the problem in July when he announced an algorithm change meant to make users see more replies from their mutuals. He said the relevant data had been missing from the algorithm, making friends appear less often and turning replies into “a battleground with people you don't recognize.” Users were understandably relieved that X had fixed it. But the episode also makes the dependency clear: one company had quietly reshaped who people encountered in conversation, and one internal change could restore it. Users had no way to inspect that decision, change it for themselves, or take their social graph to a service with different rules. The encouraging part is that competition can improve the incumbent, too. On July 15, Elon Musk said X would open-source its entire codebase after a security review and invite independent reviewers to verify that the published code is what the service actually runs. That would be a meaningful move toward transparency, even though open-sourcing X would not by itself make its social graph portable or its governance decentralized. It is also a reminder that this is not a foreign idea imposed on Twitter from outside. Bluesky began as a Twitter-funded project under Jack Dorsey, intended to develop an open social protocol that Twitter could eventually adopt. It later became an independent company, and X under Musk chose a different path. Farcaster, Bluesky, and other open-social projects now give X a reason to compete on openness rather than treating its control over the graph as permanent and unquestionable. That is a good outcome whether users remain on X, move elsewhere, or use multiple networks. This is Cory Doctorow's (@doctorow) enshittification thesis: platforms begin by being useful to users, then shift value toward business customers, and finally extract from both groups once the exit costs are high enough. The exact stages vary, but the incentive structure does not. When the platform owns the identity, the graph, and the distribution, it eventually has the power to turn the screws. Open social protocols are an attempt to change those exit costs. A user should be able to switch social-media apps without abandoning an identity. A developer should be able to build on the same graph without asking a central platform for permission. A community should not have to rebuild its relationships from zero because a company changed its strategy, sold itself, or decided a topic was no longer welcome. This is not a guarantee that every decentralized social protocol will succeed. It is an argument that the underlying capability is worth having. We already understand this principle in finance. Nobody serious says that a bank's mobile app is enough because it works today. We want the ability to hold assets ourselves, switch providers, verify settlement, and build new services without being locked into a single company's permission system. Social identity and social graphs deserve the same scrutiny. The Incumbent's Distribution Problem There is also a mundane reason many crypto opinion leaders are uninterested in decentralized social: they already have distribution. Large accounts on X did not get there by accident. They made early moves, did the work, built audiences, and accumulated social capital over years. Asking them to establish a presence on Farcaster or Bluesky can feel like asking them to start over. From their perspective, it may be negative expected value. That is rational behavior. It is not a universal argument against open social. For a new builder, an unknown writer, or someone outside the existing crypto celebrity hierarchy, a less entrenched network can be an opportunity. There are Farcaster power users who were never major accounts on Twitter. They got traction because an open, crypto-native social environment gave them a place to ship, experiment, meet collaborators, and reach early users without competing directly with the accumulated reach of every legacy influencer. That is what an open ecosystem should do. It should lower the cost of participation for people who have not already won the distribution lottery. The appropriate response is not to demand that every major account abandon X tomorrow. It is to ensure that an alternative can exist, improve, and be ready when the tradeoffs of the incumbent model become unbearable. What Actually Failed The failure of content coins should produce a better lesson: do not confuse a platform with one monetization meta. The most useful parts of crypto-native social are often much less dramatic than a creator coin launch. Wallet-native identity. Signed messages. Tipping. Subscriptions. Portable reputation. Group chats with agents and applications. Mini apps that let a builder get feedback and transact with users in the same place. An onchain record that can be used by multiple clients rather than trapped inside one company's product. These are not all equally mature, and some are easy to over-monetize. Farcaster has had its share of spam, token speculation, and incentives that distorted behavior. It remains centralized in important ways, including a Snapchain with only a handful of block-producing nodes. Bluesky has different tradeoffs, including a centralized identity layer. Neither system has finished the decentralization project. But imperfection is not an argument for surrender. Twitter has account takeovers with opaque explanations, unilateral policy changes, and a social graph held by one company. Comparing the incompleteness of new open systems to the ideal of perfect decentralization misses the actual alternative. The relevant question is not whether Farcaster or Bluesky is perfect. It is whether they move the internet toward a world in which users have more control, more portability, and more meaningful options than they do now. They do. A Better Standard for Web3 Crypto should not be reduced to an elaborate settlement layer for trading. Finance is one of its clearest and most consequential applications, but it is not the only domain where ownership and credible neutrality matter. The broader promise is that people can read, write, own, and transact without having every important relationship mediated by a platform that can change the rules unilaterally. Signed posts are not as important as signed transactions in every context, but the same primitive matters when a person's identity, work, reputation, and audience are at stake. That does not mean every interaction needs a token. It does not mean every creator should issue a coin. It certainly does not mean speculation should become the public face of every social product. The goal is not to turn friendship into a market. The goal is to give people choices before the platform enclosure becomes total. DeFi can change the world. But it cannot by itself prevent the internet from becoming a set of company towns. It cannot give creators a portable audience. It cannot make an account less vulnerable to a centralized platform's mistakes, incentives, or whims. For that, we need decentralized social. We need better protocols, more decentralized implementations, and social products that lead with connection rather than financial extraction. We need builders willing to keep making small, reversible bets rather than declaring an entire category dead because one speculative expression of it failed. Enshittification still matters. And so does building an exit.

## Publication Information

- [DFern Longform](https://paragraph.com/@dfern.eth/): Publication homepage
- [All Posts](https://paragraph.com/@dfern.eth/): More posts from this publication
- [RSS Feed](https://api.paragraph.com/blogs/rss/@dfern.eth): Subscribe to updates