The Fediverse is Not the Way Forward
If I do say so myself, I have done a commendable job of tearing myself away from the clutches of Big Tech and social media. Facebook I maintain only for its messenger. Reddit I only use to find information unavailable elsewhere. Everything else I never really used.
But occasionally I'll feel a tug toward a more social web-browsing experience. It would be rather nice to purposefully immerse myself in the Discourse™ rather than relegating my online reading to a simple, lonely RSS feed (albeit one that I created specifically to avoid the Discourse™). But I stepped away from Facebook and Reddit for some very good reasons, which I'll outline below. Isn't there another way to engage with social media that isn't so terrible?
Why, yes! At least, so claim proponents of the Fediverse, who, as I understand it, likely have an appreciable overlap with people who care enough about the Smallweb to unearth no-name sites like this humble blog. I recall dipping my toes into the Fediverse several years ago, but not sticking around. Recently I decided to drop in again to see how it's evolved, and whether it could prove a worthy investment of time for someone like me, who has the wherewithal and tech savviness to navigate experiences less tailored to the lowest common denominator but decidedly does not have the free time or energy to commit himself to a half-baked, user-hostile, and frustrating labyrinth of design. So where does the Fediverse fall?
To give some order to these thoughts, I'll first explain the things that drove me away from traditional social media, followed by the understanding I'd gleaned about how the Fediverse ostensibly remedies those things. Finally, I'll recount my research about whether it lives up to its hype before laying out my verdict and detailing my future plans.
Big Tech Hates Your Guts
At some point a few months ago I came to realize that browsing Facebook no longer provided me with any benefit. In fact, my moods were always decidedly more sour after scrolling than before. Facebook was supposed to be my way of staying connected with acquaintances and friends, but had long since devolved into an endless torrent of ads, AI slop, and posts from pages and people I'd never heard of before and certainly had never subscribed to, but had been foisted onto my wall regardless. As well, and far more discouragingly, it became clear that staying in touch with people I actually know, or knew, provided far less value to me than I had expected. Everything had become so stressful, and almost everything was centered around politics. Everything was so polarized that my feed could be an echo chamber or a battleground, but nothing in between. That is not conducive to healthy interaction, and it's especially not healthy for relationships I hope to maintain.
Reddit, meanwhile, retained its utility, but only provided that I used it as a visitor rather than as a participant. As a visitor, it is among the final bastions of the Internet that you can be sure is at least mostly composed of human-generated content. (That may change in time as AI slop continues its loathsome permeation, however.) Since Reddit itself is entirely text-based, it remains accessible and efficient, and appending site:reddit.com is still the best way to make Google search queries return anything approaching useful. As a participant, though, the experience leaves much to be desired. The oft-denigrated Reddit "hive mind" is an effective, albeit rather oversimplified, way of conceptualizing it. If your post or comment reaches a score of 0, it is incredibly likely that it will be dogpiled upon by people who downvote it just because that's what everyone else is doing. If a Reddit score ever was a decent measure of value, it certainly is not now, nor has it been for many years. Comments that are censored by virtue of their low score are usually just as constructive, helpful, or relevant as those that are highly upvoted; to believe otherwise indicates, in my mind, that one has fallen prey to some heavy confirmation bias. Even if you fall into lockstep with a particular subreddit's culture and behavioral expectations, Reddit moderators are infamous for abusing their power. You can be banned for no apparent reason and left with no recourse or appeal, or even an explanation after the fact for what you did to invoke the mods' ire. Strangely, Reddit as a whole has a very bipolar cultural response to mod- and downvote-related criticisms; if you share these criticisms, sometimes you'll be met with commiseration, while other times people will use your experience as evidence that you're secretly a troublemaker. Given the sheer volume of subreddits, it's possible to leave behind the ones with particularly nasty mods, but you can never be sure when some clique of powermods has left you banned without notice from subreddits you hadn't even visited. (In preparation for this article, I submitted a post to a subreddit about alternative social media platforms asking for clarification on my understanding of the Fediverse. I found out that I had been shadow-banned from the subreddit, and the mods ignored my request for an explanation as to why. The ordeal was, at least, a perfect encapsulation of many of the reasons Reddit is a dumpster fire.)
Though I don't have any direct experience using Twitter, it's such a huge part of modern communication that I'm aware of how it is, and how it used to be. While before it could be characterized as mostly well-adjusted people tip-toeing around a nest of toxic Left-wingers who used it as a cancel culture cudgel, since Musks's acquisition it has become a place of mostly well-adjusted people tip-toeing around a nest of toxic Right-wingers, interspersed with actual white supremacists, who primarily use it as the setting for a big, self-congratulatory circle-jerk.
Tiktok is just a hellscape of AI slop and corporate algorithmic manipulation perfectly crafted to melt the attention spans of anyone who uses it.
I don't know about the others. I think people still use Snapchat. Do they still use it to send nudes to each other? And people use Instagram. Maybe Instagram is nice these days.
Anyway, we can distill the fundamental problems with traditional social media in this way:
First, they are completely centralized. On Facebook, I am beholden to the whims of men in suits who have gleefully incinerated the mental and emotional stability of billions of people for the sake of making money. On Reddit, I am not only beholden to the admins' whims, whose motivations are equal to Zuckerberg's, but also to the whims of each subreddit's mods, who are often emotionally dysregulated and not fit for managing a lemonade stand, let alone a community of people. This means that any adverse consequences doled out to you cannot be subverted, and you have no other option but to dance to their tunes.
Second, they have cultures of toxicity. Facebook is purposefully designed to make you upset, and its users play their part in exacerbating the issue. Reddit's upvote/downvote system results in each community nurturing its own hive mind, all of which are unique but uniformly deplorable, like so many tiny, yellow snowflakes.
Third, they are algorithmic. Long gone are the days when my Facebook wall was just a series of posts by the people I'd told Facebook I'm interested in hearing from. Reddit suffers less from this if you ignore the home page, but the upvote/downvote system has the tendency to elevate or censor content not by its quality, usefulness, or truthfulness, but by how palatable its users find it.
Fourth, they are opaque. Facebook and Reddit are not going to openly admit to you how much of your data they have collected without your consent, share with you the profits they make by selling or using this data, or allow you to delete your data from their servers.
Fifth, they are beset by interminable streams of AI slop. On Facebook it's usually because people can't tell what's real, and on Reddit it's usually because people can't be bothered to articulate their own thoughts and rely instead on LLMs to do it for them.
Sixth, they benefit from the network effect. Facebook and Reddit are too big to fail, because any new upstart alternative is going to be pointless unless the majority of users make the switch simultaneously. Even the most well-designed and well-intentioned alternative is going to wither in the womb because without content, people won't come, and if people don't come, there won't be content.
The Fediverse to the Rescue?
Enter the Fediverse, whose conceits I will attempt to explain in my own words. Every site built on the Fediverse exists on top of a protocol designed to be more open and decentralized than traditional social media. Indeed, to call them "sites" is not perfectly accurate; they are more like services which provide a template by which you can make a site of your own, called an "instance." There is a service that basically copies Reddit known as Lemmy, for example. I can take Lemmy and make my own little Reddit-clone website, and then people can sign up for that instead. Or I can use Mastodon, a service that basically copies Twitter, and use it to make my own little Twitter-clone website. The main benefit is that even if you sign up for my little Reddit-clone, you can still talk to people who signed up for someone else's little Reddit-clone. The services we built them with are designed to allow each implementation to talk to each other. So even if only five people sign up for mine, they can participate in discussions happening on everyone else's, meaning the userbase of each is effectively much bigger. This means you can choose a Reddit-clone whose admins and moderators you trust to do a good job, instead of there existing one set of admins and mods who you simply have to put up with. This capability of different sites to talk to each other is called "federation." As indicated, there are different Fediverse versions of several social media sites, including Twitter, Reddit, Facebook, Instagram, and YouTube. Thanks to federation, these services can all talk to each other. "Your data belongs to you" is a common refrain by proponents of the Fediverse, referencing the fact that you can move your account from my Reddit-clone to someone else's if you end up not liking how I run things. Plus, you can see posts from Twitter-clones on a Reddit-clone, for example, meaning that the stream of content is even bigger.
To summarize, the Fediverse is sold to newbies as signing up for an open-source alternative to traditional social media, where you are protected against bad actors in positions of power, where you can participate in discussions freely no matter which implementation of the service you sign up for, and where your data (a vague term that seems to just amount to your followers and your list of who you follow) belongs to you because you can move it freely between sites and services. As a decently tech-savvy person, this was my understanding after researching it for some time, and this description seems to check all of the boxes for me.
In practice and with more thorough research, however, it became clear that things were not as sunny as I'd hoped.
Promises Unfulfilled
It turns out that there are some pretty hefty asterisks to the idea that your account on the Fediverse is portable between instances and agnostic regarding services. It's true that on Mastodon, the Twitter-esque platform, you can migrate your account between instances. However, this requires logging into your account, so migrating it after having been banned from your current instance (i.e., the time you'd most likely find yourself wanting to do so) apparently requires cooperation from the instance's admins, who have to be gracious enough to allow you to do it. If they refuse, your data and followers are gone forever, just like with traditional social media. Furthermore, this is a feature of Mastodon specifically; you can't migrate your Mastodon account to an instance of Lemmy.
This is a problem that compounds the more you think about it. The Fediverse has developed a reputation for its horrible onboarding process. You can't just tell someone to "sign up for Mastodon," because you'd need to help them figure out which instance to sign up for. Many of its advocates insist that this decision paralysis is foolish, because it's the Fediverse! The instance you join is essentially arbitrary, as you can take your data with you if you decide to go to another! Were accounts truly as portable as I'd been led to believe I'd be inclined to agree. The first choice of instance would be arbitrary and inconsequential, like picking which bus stop to go to first when the bus can take you to any of them whenever you want later. But absent this level of portability, the first choice of instance honestly does matter, because there isn't actually a guarantee that you'll be able to rescue your data should the need arise.
If you get past the initial cliff of indecision and manage to pick an instance (despite likely not even fully understanding, at this point in the process, what all of these terms even mean), you could just hope for the best and plan to stay where you are. It is true that most people probably won't have a need to switch instances very often, if ever. But even if you don't, there are consequences to which instance you pick. Fediverse services like to paint themselves as a place where you get to take control of the algorithm, where the content you see is up to you, not the whims of a profit-seeking corporation. This is also not true. Once you've set up your account in an instance of Mastodon, for example, you'll find that you don't get to control what you see. The system of federation, that thing that lets instances talk to each other, is not under your control, but the control of the instance's admins. So if they decide they want to "defederate" from another instance, you aren't allowed to see posts from that instance. Your feed is by no means under your control; your experience is still shaped against your will by whichever digital landlord you happened to pick from the list. This became especially frustrating when I read that it's not uncommon for instances not only to defederate from others based on the ideological differences between admins, but to defederate from unrelated third parties because they choose to federate with people the admins don't like. In other words, X may block Y because they don't like Y, but X may also block Z because Z doesn't block Y too. When you dig into it, the promise of an open, decentralized platform where you take back control of your experience just turns into a giant puzzle, where you have to figure out not only which instance has a culture you like but also how much of other instances' content the admins will prohibit you from seeing. It should not have to be necessary to cross-reference instances to make sure that the one I choose will allow me to see what I want, and also will allow other people to see me if they want.
Things don't get much better if you find an instance whose policies you're satisfied with. Recall that one of the enormous roadblocks preventing new social media platforms from gaining ground is the network effect, or the fact that unless everyone moves at the same time, the new platform will be empty, and therefore will never persuade people to move in the first place. The promise of federation is ostensibly supposed to remedy this. Rather than being stuck watching the tumbleweeds go by in an empty site, every instance mutually benefits from the proliferation of each other's content. And what's more, your instance of Mastodon benefits from my instance of Lemmy, so all of these plucky upstarts grow at the same rate and reinforce each other's growth. You'll never have to babysit an empty feed because there's going to be activity somewhere in the Fediverse.
This alone would likely be enough to motivate me to overlook the broken promise of "owning my data" and being able to "move instances whenever I want," but even this is, at most, a half-truth. Reports of broken, laggy, or incomplete federation are very common. You may see a post that originated from another instance, but its attendant data, such as reactions and comments, may come in full, or only partially, or not at all. You'll never know if you're engaging with the complete chain of interactions and replies, or if there's content that you're missing out on. You'll encounter comments which act as responses to other comments further up the chain which your instance never showed you. If you make a post, the discussion under it is blown apart and fragmented beyond recognition as everyone on other instances only gets a partial picture of the full conversation. Rather than federation allowing me to share in the discourse happening across the Fediverse, it acts more like a big fan set up in the post office, with letters and postcards flying around haphazardly with no way to organize them or know whether I've collected enough mail to recreate a full dialogue. If the promise of federation alone were enough to keep me invested, the sorry state of it is enough to turn me off of the idea completely. I've seen people recommending browser extensions to ameliorate incomplete federation, which themselves require multiple steps to make sure you see other instances' content.
If it's even worth mentioning at this point, the promise that different platforms within the Fediverse can see each other's content is also not exactly true. Some of the platforms, like Lemmy and Mastodon, have partial compatibility, requiring illustrated guides showing you how to navigate their interaction in the software. Most of them, though, can't actually see each other. There are apparently third-party apps that allow you to use one platform but not another, and other apps that allow you to use both, and still other which allow you to use yet another combination. The fragmentation of the Fediverse is built in, as we've seen, but doesn't seem to be fixed very well with third-party software.
By no means do I expect perfect, flawless software from volunteers and passion projects, but from where I'm standing, the Fediverse looks not slightly suboptimal, but downright incomplete. Federation with results this inconsistent and severe is not anywhere near ready for use in a mainstream project, and there is absolutely no chance that the average user is going to find it acceptable, assuming they bothered to jump through all the hoops required to get to a point where they could notice the fragmentation of content in the first place. Without exception, all of the big promises and points of hype that surround the Fediverse turn out to be completely unfulfilled when put into practice. You don't own your data. You can't bring it with you wherever and whenever you want. You aren't in control of your own feed. You can't reliably see content from other instances. You can't consistently see content from other platforms. For all of the effort and research required to learn what all of the relevant terms mean, how the Fediverse works, how best to pick an instance, and how to populate your feed, you don't get any commensurate benefits. When you're hooked by the promise of decentralization, you're forced instead to just choose from eye-glazing lists of smaller, centralized servers. When you're excited by the prospect of participating in discussions from across the Fediverse, you're forced to wonder precisely how chopped up what you've been served truly is.
Ways Forward
In my experience while researching for this article, many people within the Fediverse community harbor this vexingly sanctimonious attitude about these issues, like wanting a streamlined application and to be the master of your own social media feed is whiny and entitled. Why don't you just go host your own instance then? and If you want someone to hold your hand, go back to Facebook! are common retorts to legitimate criticism. This basically amounts to giving up on the idea of making onboarding and everyday use accessible for newer, less tech-savvy users. "Your data is yours!" and "You can move accounts whenever you want!" and "You can see content from other instances thanks to federation!" are just a hair above blatant lies from what I can tell.
So do I just give up on the idea of decentralized social media again for a few more years before popping back in to reassess? Quite likely, but I do foresee four other options that might be worth exploring.
First, though I lack the technical expertise required to self-host my own instance, I may rent a VPS with some federated platform pre-installed. This would allow me to get a step closer to actually owning my data, and would allow me to decide for myself what content I see, rather than letting some instance admins infantilize me. The downside remains inherent in the protocol that everyone else might be forcibly prohibited from seeing what I post if some power-hungry instance mods don't like me, and though that is far beyond my control, I've seen pretty consistent reports that people who run Mastodon instances tend to be the sort of people who interpret mildly diverging perspectives as tantamount to Nazism (along with the sort of Kafka-esque "defending yourself is proof of your guilt" counterarguments that Redditors frequently employ).
Second, I could venture into the realm of explicitly right-wing Fediverse instances, which might between themselves maintain a big enough userbase to compensate for almost assuredly being defederated by all the big Fediverse players and which I tentatively expect to be less happy with the banhammer, given that until the Charlie Kirk assassination fallout, cancel culture was largely a tool of the Left. But then I'd be signing up to surround myself with exactly the sort of people who make my time on Facebook so miserable. I think no social media at all would be better than social media filled with Trump-glazing, 2020 election denialism, varyingly overt degrees of white supremacy, hostility toward workers' rights, and scoffing at the idea of doing things to help the poor. I don't want an echo chamber that just reaffirms my own beliefs, but would it be any better to step into an echo chamber that just reaffirms the diametric opposite of my beliefs? I don't suppose there exists any appreciable counterculture of progressively minded users on a platform like Truth Social, and I don't have the fortitude to start one. Perhaps this one isn't worth considering after all...
Third, I haven't even mentioned yet that there exists an entirely distinct network of federated platforms, the kind that BlueSky runs on, none of which is natively compatible with any of the platforms I've mentioned thus far. I could delve more into the protocol behind BlueSky and its lesser, related cousins. I've read that this protocol is designed in such a way that for all intents and purposes, it is completely centralized around BlueSky's own servers even if technically it doesn't have to be so. This wouldn't solve the problem of centralization. I have read up on some Reddit alternatives on that protocol, which might be fun, but I never really see any talk of interplatform federation of the kind promised by the Fediverse. But then again, I have read that bridging between this protocol and the Fediverse isn't terribly complex, so perhaps I could get a Mastodon VPS and bridge it with BlueSky's protocol. I don't know if such an effort, not to mention monetary expense, would be worth it.
Fourth, there also exists another, completely different protocol that caught my attention which seems a good deal closer to what I'm actually looking for in a decentralized social media solution. As I understand it, Nostr is a protocol that doesn't involve you choosing an instance on which to make an account. Rather, you generate a pair of long alphanumeric strings, called keys, one of which you keep secret and one of which you publicly attach to your profile. As I understand it, the app you use to access a Nostr-based platform makes sure your private key matches your public key and thereby allows you to make posts with that profile. The primary effect of this is that your account is truly portable, because you don't have an account on a server; you just have a pair of text-based strings to verify who you are. Likewise, Nostr doesn't use instances, but "relays," which basically listen for activity using Nostr and broadcast it to any apps who are listening, such as the one you'd use to browse your feed. This doesn't really eliminate the centralization, as you could conceivably provoke the operators of the relays to refuse to proliferate your posts. But since relays are used mostly to disseminate data rather than store it and serve it to clients, people seem to imply that most relay operators are content-agnostic. Regardless, there isn't just one relay, so if one blocks you, or just shuts down, you can use the others; the only time you'd be completely censored is if all of the operators of all of the relays decided together to block you.
This is the one I've already started exploring. Currently I've just used a Nostr-based Twitter feed, but there are apparently other platforms you can use with the same account, since, again, your account is just a pair of text strings. I don't know how long I'll stick around, though. Nostr allegedly has its roots in Bitcoin enthusiasts, so the zeitgeist there largely revolves around A) circle-jerking Nostr users themselves for being so forward-thinking and B) cryptocurrencies. I've encountered a fair few libertarian cryptobros there. On the bright side, a libertarian cryptobro is far easier to stomach than someone deep in the MAGA camp, so as long as the overlap isn't terribly extreme, I won't get the shivers browsing my feed. There does exist a relay you can subscribe to that ostensibly bridges Nostr and the Fediverse, but I need to do some more research about how reliable it is.
So that's where I am at the moment. If you care to follow me on Nostr, my public key is: npub1dxdkvwnjnhrlyjvw8n4x6r4theufum98rufq2zsyx7j9h847y34qlaz933
If I have made any grievous errors with my understanding or explanation of the Fediverse, please let me know! My email address is on the homepage.