Whatever happened to our dream of decentralization?

Elon Musk is buying Twitter for $44 billion USD. Surely. Certainly. Maybe. It’s complicated. Regardless, the news have exploded in tech and mainstream news, everyone has a take. Musk announced that he would be a champion of free speech, prompting cheers from the right (who think that this means that Trump will be allowed back in), and cries from the left (who think that this means that Trump will be allowed back in). There have been calls from users and celebrities to abandon the app and move elsewhere, more on that later.

I have to admit that I have found some of the reactions to the news a bit over the top. A common thread from some critics of the sale is that Twitter is a very important platform, one that cannot be allowed to fall in the hands of a single billionaire owner. Others feel that Musk’s free speech stance will make the app something close to 4chan and 8chan, lawless spaces filled with right-wing trolls.

I’m not so sure much will change to be honest, so I’ll definitely be staying put. My dislike of  Elon Musk is on record, (and not just because he reposted a meme hours after I had posted it), but I have found some of the responses about the news to be completely overblown. The main problem with the criticism of Musk’s takeover is in the nature of Twitter itself. In terms of numbers, it is one of the smallest social media sites, but its cultural and political influence is larger than sheer numbers, as it is disproportionally inhabited by politicians, journalists, celebrities, and academics, and therefore it has a reputation of setting the tone in many conversations. But because of the people who use it tend to be mostly left-leaning cultural “elites”, it acts as a powerful filter bubble, enhancing whatever you already believe to be true.

In the end, I don’t think that Twitter will change. You are swapping one crypto-obsessed tech-bro libertarian with another. And sure, having platforms owned by billionaires is not the best thing, but that is the norm nowadays. Musk has done quite well from using Twitter to boost his stock, and he truly enjoys his impish shitposting memelord persona in the platform. I don’t think he will toy too much with what makes Twitter great (or a doom-scrolling hell-site for others). On the contrary, we could see some positive changes as the platform stops relying on enhancing shareholder value at all times, perhaps opening a few tools even.

But regardless of what happens, many Twitter inhabitants have decided to leave the platform, which brings me to the main topic of this blog post: decentralization.

So if Twitter is no longer viable for many people, what is the option? Ideally, those who object to Twitter also distrust other centralised, billionaire-owned platforms such as Facebook. Reddit is not really an equivalent, it’s more of a forum than a social media, and TikTok is for kids (honestly, I feel ancient anytime I log in there). There is currently only one alternative, and it is Mastodon, an open source self-hosted social network.

On paper, Mastodon sounds like a dream social network for people like me. It’s the very definition of decentralization, anyone can download the software and run their own instance, or users can join an existing server, each with its own moderation rules, code of conduct, and privacy settings. Instances are independent of each other, but you can send messages and communicate with users across different instances. The distribution of users ranges from instances of one, to more popular nodes with hundreds of thousands of users, and everything in between.You can join an open node, or you can request to join a restricted one.

Those who use Mastodon praise it for being smaller communities of like-minded people that lack the chaos and vitriol of places like Twitter.

However, Mastodon is a good example of the limits of decentralization. In an interesting Twitter thread, user @atomicthumbs listed a good number of problems that come with having a decentralized network. Because each node is run by a hobbyist, users of that instance are at the mercy of the system administrator, and their entire account can disappear if the admins decide to shut it down for whatever reason, and when it’s gone, it’s gone. More worryingly, the admins can see all the personal messages of the users, as well as all of the incoming messages from other instances to users in your node. This is a privacy nightmare.

There are also several usability issues, which I experienced first-hand when I tried to join the Fediverse. At first, I did not understand the instance concept, and I thought that if I created an account with one server it would be replicated elsewhere, but that wasn’t the case and I ended up with 3 technollama accounts in different instances. And yes, people can join with your same handle in other instances, so there could be people impersonating you in other servers. The solution is for users to run their own node, but how sustainable is that for a social network? Can we exist in an environment where everyone is supposed to be a computer security expert all of a sudden?

Social media is just an example of the problems that we have when we try to decentralise certain technologies. Another example can be found in cryptocurrencies and the blockchain space.

One of the biggest selling points of blockchain technology has been that of decentralization, an antidote to the centralised tech giants, greedy corporations, bloated financial institutions, and stagnant governmental institutions of the world. Here we have an open source, cryptographic, public, transparent, decentralized, distributed ledger, there is no central authority, it is governed by maths and trustless consensus mechanisms.

But these decentralized credentials tend not to survive scrutiny.  The first problem is that decentralization comes at a price. Centralized systems tend to be more efficient because the costs of running the network are paid by the central authority, while a system that has no central control must find ways of paying the bills. In many decentralized systems this is done by hobbyists, dedicated people willing to pay for the resources themselves, as is the case with Mastodon. Other decentralized services can operate as non-profits thanks to charitable donations. But in crypto the decentralization is paid for by environmentally-damaging mining, or by bringing in capital from wealthy investors, hence becoming more centralized in the process.

One thing that I have noticed in Web3 is that they tend to be expensive and slow, the more popular the system becomes, the slower it gets. This is built-in the system, as the decentralization protocol discourages anyone from using too many resources, or from gaining control. But this means that blockchains are not scalable, and are doomed to be extremely poor alternatives to centralized systems, thus cementing the centralization of everything.

The solution is actually to centralize more. One way around the scalability problem is to build secondary networks on top of the slow decentralized blockchain, but these act actually as central nodes, sort of defeating the purpose. Something else that can happen is that decentralized systems start accumulating nodes, which make them no different to Web2 equivalents. OpenSea has now a 90% share of the NFT market, something that approaches Google-level of centrality.

Concluding

Perhaps we are doomed to rely on centralized systems. From social media to finance, decentralized solutions can only succeed if they are small. If they start scaling up, then they stop being decentralized, just like crypto and blockchains, which are decentralized in name only, and display astounding levels of centralization and accumulation of wealth in very few hands.

But in the end that is the whole point of decentralization. We either choose to operate in centralized systems, or we break everything up. Go small at all levels.

I have to admit that I do not have the energy to move to Mastodon, and I am extremely sceptical of crypto as a decentralized system, but I haven’t given up on decentralization yet. My own contribution to it is this blog, which is a stand-alone, self-hosted, and self-financed experiment. A remnant of an Internet long gone, a relic, an elegant weapon for a more civilized time.

I need to keep the dream alive somehow.