Alright, I just finished watching Vitalik’s excellent discussion and that pushed me to finally rant about a topic that I’m interested in – decentralized social media. So, here’s what a decentralized social media protocol looks like, at the top of my mind. I know all of this is going to sound vague – because it is.
The most important thing is to build a protocol, not a product. This needs to be a common protocol that will allow developers to build social networking apps on top of, can be used by multiple types of frontends, varying architectures, revenue models and incentive mechanisms. The key goal of the protocol will be to coordinate content, financials and identity.
Think of it this way: Facebook, Twitter, Instagram share the same content, identity, and financial infrastructure, which are interoperable with one another – yet they express the content in unique ways.
From the end user’s perspective – you just Sign in with Ethereum wherever you want.
Blockchains are ill-suited for content availability. Currently, you must have a very, very centralized blockchain that’ll store all content, and even then, there are some hard limits to scalability. With a very, very centralized network, you’re compromising the chief advantage of blockchains for content anyway – liveness and censorship resistance. Moreover, it’s quite a waste of space for an entire blockchain network where every node has to store your worthless shitposts.
We need a different storage solution. I don’t know what the right technical solution looks like just yet, but projects like Swarm offer promise. It could also be something like BitTorrent with erasure coding techniques. This way, the promise is that a network of very many nodes only need to hold a small amount of data, but together, there’ll be an entire social network with petabytes of content. The economic incentives will be key, of course.
I expect content in the future to diversify from text, to images, to video, to interactive real-time graphics. How a content infrastructure can scale with this is an open question. It’s important for this protocol to offer options. It could be that a certain app built on it wishes to use a centralized database, or their own data peering systems or committee. They should absolutely be able to. There should also be an option for those who want to pay for it to commit content (potentially private) to a blockchain, even if it’s expensive.
Moreso than censorship resistance, the greatest trick a blockchain-based social network can pull off is combining the financial world with social. More on this later.
Architecturally, this one’s simpler – you need full data availability guarantees for financial applications, so a blockchain is a requirement. Currently, the best option would be a volition-like zkVM setup. It could be built on top of, say, StarkNet; or its own zkVM built on top of Ethereum. Different applications and users will use different data availabilty solutions as per their security/affordability needs – though the default will be Ethereum data shards.
Identity, attestation economy, and reputation systems
The key infrastructure that’ll enable all of this is systems like ENS, Proof of Humanity, POAPs, etc. It is crucial that there’s a solution for identity and preventing sybil attacks, while also inviting anonymous and alternative identities. This is definitely a very nascent field, and it’s hard to say how all of this will be solved. Autonomous reputation systems will need to be innovated upon – more on this later. For now, though, we have to ensure that the networks are not attractive to sybil attacks – e.g. please don’t make tokens farmable.
What about advertising, then? Decentralized advertising is still an unsolved problem. Solutions can be built at the protocol level, but they can be overridden by the applications and frontends built on top. So, the idea is to let the applications extract advertising value.
How is this any different from the traditional world? The answer is simply – competition. The underlying protocol is a decentralized public good, and applications have to compete to be the best solution built on it. Friendlier and more interactive advertising forms are also due. Today, Facebook can get away with all sorts of anti-consumer behaviour because they own all your data, content, identity and have unreplicable network effects. A decentralized protocol like so greatly levels the playing field, as the data, content and identity now rests on a decentralized layer. It’s true that a certain application could build network effects, but the opportunity for disruption is always open.
Beyond this, I have many thoughts on how a particular application can take advantage of smart contracts for a more efficient advertising and attention economy, but that’s for later.
It’s not just about advertising, of course, gamification will play a vital role for certain applications. Reddit has experimented with Coins, rewards, and now, Community Points. We can build on top of this. Combining this with DeFi is when things get really interesting, and beyond anything a social network can do today. The metaverse is real, and a decentralized social networking protocol will be at the heart of it. One of the great challenges here is mitigating the toxicity inherent to financializing social – more thoughts on this sometime else.
So, I’m quite biased, as much of my crypto time in 2016-17 was spent working on the Curie curation DAO on Steem. It continues to live on today, on the Hive fork albeit downscaled given the fate of Steem/Hive, and is constantly sorting the Hive ecosystem with high quality content. Formed in July 2016, right after the DAO fork, I’d argue it’s the first successful DAO in the blockchain space. I believe in decentralized incentivized curation systems significantly improving the social media experience for everyone, where everyone is a potential curator and a consumer. Curation systems will tie in directly to reputation systems mentioned above. I also believe this will be a hybrid system with algorithms and human curation mixing things up over time. Finally, the applications built on the top of the protocol will be incentivized to innovate and offer the best curation and content discovery possible.
Weaning users off Twitter, Facebook, Instagram, and other entrenched players will be a gargantuan task. But this is how we can get started. An application can build an overlay solution, which gives people a glimpse of what is not possible with centralized social network. For example, the user can continue to post on Twitter, but a mobile app or browser extension will link and expose them to gamification, better incentivization, better content discovery etc. The hope is that over time, they’ll realize that it’s Twitter that’s dispensable, not the metaverse.
With a zkVM with precompiles, privacy using zk-SNARKs becomes possible. This is the first time true, decentralized privacy becomes possible. It’ll cost more, and the user might have to pay for it, but it’s an option that will exist. Zero-knowledge proofs enable a new paradigm of granular communication of information with varying degrees of privacy (e.g. I can tell you specific things without you having to know the full picture).
As you can tell, I have a lot of thoughts on the matter, but unfortunately, I have limited bandwidth and time. For now, I’ll focus on discussing rollups till the world has awoken to the wonders of it. Hopefully, by then someone will figure it out!
Cross-posted to my blog: https://polynya.medium.com/decentralized-social-networking-b21a94180477?source=your_stories_page————————————-