Building Decentralized Social Networks of Tomorrow
Concluding the series of articles about the shortcomings of DPoS and LPoS algorithms, Vladimir Popov talked about the decentralized social networks and the internet of the future where websites can’t be shut down and access to information can’t be restricted. This is an adaptation of the original Russian-language article Vladimir wrote exclusively for ForkLog.
Decentralized Social Networks Today
There are several solutions already up and running:
- Golos (previously a Steem fork, a Dapp on CyberWay blockchain now).
- dTube (a mix of an IPFS and Steem for videos. The closest comparison is Livepeer but more similar to stock).
- Hive, a Steem fork that was created because of the situation with Tron.
- Commun, a decentralized social network on CyberWay blockchain.
- VIZ, a DAO for digitizing the social capital of projects like YouTube channels, Telegram bots, etc.
- Memo, a rudimentary online notepad on Bitcoin Cash.
There’s even an entire protocol for building social networks called ContentsProtocol, as well as services like cointigo.io aimed at monetizing accounts in traditional social networks. A service called Bitrad.io monetizes radio streams.
Voice and Block.one, which are currently in development, introduce payments for verification and other bonuses. There’s a project called Twister (a decentralized Twitter) that implements the usual functions of traditional social networks. As an alternative to Tinder, there’s Legalfling, for Wikipedia there’s Everipedia. There are also decentralized projects venturing into VR space, Decentraland being a prominent example here.
Although, despite the innovation, each of these projects tries to put a DDS (decentralized and/or distributed system) under old ideas and architecture principles. There are downsides to this approach.
A decentralized social network (DSS) can take the following forms:
- Dapp, a decentralized application.
- DAO, a decentralized autonomous organization.
First of all, there’s a misconception that DAO is narrower than DeFi, which isn’t the case since you can digitize not only the financial capital but also social and other forms of capital. Next, the difference between Dapp and DAO can be conditional in terms of functions, implementation complexity, etc.
General Solutions for DSNs
The more storage options, the better. People are too focused on IPFS. It is an effective solution, but there are other p2p storage options out there.
Looking at the test network, Filcoin is already something very different from IPFS. The same goes for Sia, Storj, and others. And this is not the last step. Moreover, there are older networks like I2P and Tor, as well as torrents.
A DSN has to be not only massive, built on trust and cryptography, and open, but also fair. Nobody calls for digital communism, but at least the active and useful members of the community should be rewarded. The risk of a hostile takeover via staking and validation should be minimized.
There’s one principle to keep in mind to achieve full decentralization: each stage of building the DSN should have its own level of decentralization (domain, hosting, etc.). This is impossible without a large number of participants like miners, full nodes, users, developers, etc.
The next step is equality. For a DSN it is important to have humans and robots equal in terms of transactions but clearly distinct.
One of the ways to solve this problem can be a global reputation system: everybody joins any network with zero reputation and then proves themselves via transactions. If a user can quickly solve complex mathematical problems it doesn’t matter that much if that’s a genius human or a new algorithm. This way, we don’t wonder about the sentience of a machine but focus on its usefulness. That’s exactly what Facebook and Telegram do. Sure, there are certain limitations when it comes to registering, but there’s no shortage of fake accounts nonetheless.
The first signs of such equality are already present:
- A virtual influencer Lilmiquela has over 2 million subscribers on Instagram.
- Hatsune Miku is doing well since 2007.
- Participants in Second Life and Decentraland cost real money.
The edge between the virtual world and the real one disappears, which is a prerequisite for Web 3.0 and DSNs.
Short Retrospective and Next Steps
There are several prominent examples of implemented DSN prototypes:
- Steem, 2016, Steem blockchain.
- Bitwalking, 2016, Dapp.
- Golos, 2016, Dapp.
- Livepeer, 2018, Dapp.
- SportCoin (iTerra), 2018, Dapp (DDS).
- VIZ, 2018, DAO.
- Commun, 2019, Dapp.
- dTube, 2019, Dapp.
There are also messengers like Adamant, Bitmessage, Crypviser, Dust, E-chat, Eco, MyHush, Sence, Status, Tox, Will Channels, and Keybase, all of which have even smaller userbases.
It is important to look back at history to understand things better and figure out if there is progress.
An obvious example is Facebook. It became a thing in October 2003 and in 2005 they bought facobook.com domain name for $200,000. Such a sum would be considered normal for a pre-ICO, which leads to a suggestion that the amount of money invested in p2p projects is far less than in most VC startups for similar timeframes. Facebook itself got $240 million back in 2007, about four years after launch.
This is where some people argue that traditional social networks took 3–4 years to develop into something big, while DSNs appear to need much more. But back in the early years of traditional networks, there was no competition. Now the competition is off the charts. We see the rise of video services like TikTok and messengers like Telegram and Snapchat, not traditional social networks.
One of the biggest challenges before DSNs is to prove their uniqueness, which is base on censorship resistance, activity monetization, and integration between online and offline parts of people’s lives. By now, none of the DSNs reviewed have managed to utilize these qualities to make a breakthrough, but the 2020 crisis is an opportunity to find new audiences and investors.
Summing up, there are several key takeaways about DSNs:
- Decentralization achieved in PoW and PoS systems is more of a transitional form.
- When building a DSN, it should be considered that each type of participant (validator, user, miner, etc.) has to be a part of a group with a minimal distribution of rights and abilities (3–10 people); the general structure should be expandable (be positively inflatable). This will allow exploiting the benefits of the network effect and let the community control all the links of the network
- Equality between humans and artificial agents is inevitable in the Web 3.0 paradigm. Cyborgs, bots, neural networks, IoT devices, smartphones, computers, digital avatars, and others will be equal to human users. To do that DSNs have to move from subject-based identification to transaction-based. Each action or inaction should be evaluated in terms of the benefits it brings to the network or its part.
- Different types of attacks related to decentralization, such as Sybill attacks, shouldn’t be out of the question. There have to be protection mechanisms against those. An attack should not be economically viable. Only then additional levels of protection like the Winkle algorithm will be effective. The complicated part is that unlike traditional social networks, attackers can be after full destruction of the network. For this reason, self-governance systems like sharding, PoI and PoA, Fair DPoS, mobile mining, or others are required.
- Self-completion is a required consequence of a DDS architecture, so it should be used in DSNs. Otherwise, its benefits become downsides. This is more important than TPS or protection from bots.
- The levels of protection themselves should be built on different bases. Instead of punishments, there should be rewards that depend on interests and usefulness. Instead of personal data, there should be a value of reputation.
There are other aspects to DSNs that have been reviewed multiple times already. The pieces in this series aim to explore less beaten parts of DSN creation and help developers, system architects, and users to see the unique benefits of DSNs instead of the obvious things they have in common with their traditional counterparts.
Subscribe to our Newsletter<
- Enlightened Despot: Is Google Fit to Tell Good from Evil?
- Mastodon Review: Federation of Bubbles
- NewTON Blockchain Launched Test Network, Building Upon TON’s Original Source Code
- Are Censorship Free Platforms Doomed to Become Den of Trolls? Gab’s Case
- TON Developers Pushed Final Code Updates, Now It’s Up to Community
- Intellectual Hygiene: Six Rules to Battle Fake News and Propaganda
- Is First Decentralized Social Network Still Relevant: Steemit Retrospective Review
- Searching for Search Engine and Post-Google Dream