In April, in the wake of Elon Musk’s move to acquire Twitter, Andi McClure left that platform for Mastodon. Andi’s tweets and retweets have been for me a continuous source of wonder and delight, so I moved after her.
Unlike Twitter, which is centralised, Mastodon is a collection of servers (instances) that communicate with each other. They use the ActivityPub protocol, which is a W3C-recommended standard for publishing and receiving all kinds of status updates. Each instance is administered separately and can decide which other instances to communicate with. What is more, the instances can run different software - not necessarily Mastodon - and still participate in the global, federated network. The fediverse (or fedi for short) refers to this global network, and also includes federated services that might be using communication protocols other than ActivityPub.
All this means that creating a “Mastodon” account requires first choosing an instance, whose domain will become part of our user handle and who will act as our window into the federated network. I’ve read in one of the joining guides (I can’t recall which one) that it’s better to not default to the biggest servers, such as mstdn.social, because there I’d miss out on the sense of community that distinguishes fediverse from big social media sites, so I chose a smaller instance with a couple of thousand accounts.
The first impression of the federated feed was a confusion. What is all this right-wing shit? Are these swastikas for real, or is it a particularly tasteless trolling? And what about all that porn - is no one moderating it? I soon discovered the ability to mute users and - particularly useful - block entire domains, servers that might be lax on moderation or attract people I’d rather not interact with. In this way I was able to largely recreate the info-bubble I have enjoyed on other platforms.
On Mastodon, this is supposed to work as follows: each instance’s admin defines the content policy; for example, they might disallow racist and sexist posts, or permit anything that the law of a specific country does not forbid. The administrators are then responsible for enforcing this policy via moderation of posts on their server, as well as not federating with instances whose members post content that violates these policies. In this way the choice of the instance is the first line of defense against offensive content.
While the fedi content distribution and moderation model makes a lot of sense, and my own experience has thus far been positive, I suspect it might not scale to the hundreds of millions of real and bot accounts that the largest social media sites have to deal with. It could also prove vulnerable to the sophisticated state-funded abuse and manipulation. Popularity, should it ever happen, might turn out to be fediverse’s undoing. But that’s fine; nothing lasts forever, and if fedi degenerates into another cesspool, something else will spring up.
For now, Mastodon is a good place to be. Largely thanks to the interactions feeling more meaningful than on the large commercial platforms, the presence there not being about promoting a persona or building a brand and monetisable following, but participating in communities with common interests and causes. Even the user interface, with content warning support for hiding posts that some readers might find upsetting, and marking media as sensitive for the same reason, conveys a sense of empathy and care about others’ feelings.
While Mastodon might be the most widely recognised, there are federated alternatives for many popular social media platforms:
|Mastodon, Pleroma, MissKey|
and many more.
Twitter, Facebook, YouTube, you name it, made us accustomed to seeing a feed generated by an algorithm whose objective is to maximise the ad revenue, not our experience. Our behaviour is tracked and analysed to help the corporate bottom line, sometimes also to influence elections or other political outcomes. Mastodon, and, by extension, fediverse, allows us to take back control of our information feeds. We see the posts from accounts we followed, in a chronological order, not some random mashup of activity we follow with promoted posts, overt ads and other noise.
In this and some other ways fedi feels like a return to the 1990s, when the early web was a collection of independent websites, some publishing useful info, some utter garbage, but without the Googles and Facebooks deciding what we see or not and tracking our every step, us trading the privacy and freedom of exploration for convenience. It’s positively exhilarating.
The sense that Big Tech has appropriated the space for digital human connection, exchange of opinions, community forming, entertainment and creativity, is widespread. One well-known movement that tries to break free from the corporate confines is web3. As far as I understand - and I am not able to make much sense out of most of web3 discourse - the focus of this effort is asset tracking and monetisation of works and services that does not rely on the centralised channels controlled by the Big Tech. In this sense, it is politically opinionated: the capital and wealth creation is its core tenet. Perhaps that’s why venture capitalists do their best to promote web3.
In contrast, fediverse does not care much about the technology, funding sources or profit: it is about the ephemeral human connection. Its politics is just about the freedom of individuals to interact and form communities without the corporate surveillance. This is a very wide umbrella that can accommodate libertarians, anarchists, communists and progressive liberals alike. It shares a lot of characteristics and ethos with the Free Software movement. Take the control over computing away from corporations and large institutions, empower individuals to use the computers in the way that serves their interests.
Fedi is more ethical than the corporate social networks as far as resource usage is concerned. The sole purpose of the network is to let people communicate with each other, and the dominant funding model, where the users and instance operators chip in to pay for running the instances, promotes the use of just enough infrastructure capacity to support this core function. The commercial networks, on the other hand, are funded by ads, which are paid for by brands that in turn get their revenue from our consumption of ever more goods and services we don’t necessarily need. The commercial network’s infrastructure must not only support the social interactions, but also ad serving and optimisation, thus consuming more resources than necessary to let people talk to each other.
The fediverse is perhaps the most exciting thing in the social use of computing I encountered over the last 15 years. Sure, there are hiccups, imperfections, abandoned instances, abuse. But compared with the curated, sanitised walled gardens offered by the Big Tech, this patch of wild growth feels like a breath of fresh air.