A quick warning about this episode of The World on a Wire Show: this time we’re talking about social computing and the way people behave on social media. We’ll be talking about some of the worst behaviors that occur on social media and how they contribute to real-world violence. I’ll try not to be too graphic about these topics, but they will come up.

They say it’s friendship. We say it’s unwaged work. With every like, chat, tag or poke our subjectivity turns them a profit. They call it sharing. We call it stealing. We’ve been bound by their terms of service far too long—it’s time for our terms.

These words are excerpted from the web site Wages For Facebook, which launched in January 2014 and at time of recording exists in its initial form, as a wall of dense, scrolling text making the case that the people Facebook calls its users are workers who generate the company’s enormous profits, and that these workers should demand wages from the company they have made so successful. It generated a lot of buzz, but right now just about everyone I know engages with Facebook daily or almost daily, giving Facebook valuable information about where they live and work and what their interests and affiliations are, and Facebook isn’t paying any of them to do it.

The World on a Wire Show, season one, episode three.

The social market

Free Software Foundation founder Richard Stallman is no longer president of the foundation for good reason, but his opposition to the privacy-violating behavior of corporate social media has proved a lasting influence, and this specific point, that Facebook uses the people it calls its users to gather personal data that forms its greatest business asset, has been borne out by the evidence. In an April 2017 interview with Linux vlogger Brian Lunduke, he put it like this:

Most people who use the internet are using disservices which basically exploit them to get their personal information. And even if you’re using a free browser to talk to them, they’re still mistreating you. Nobody uses Facebook; Facebook uses them. Facebook doesn’t have users; it has useds. And if you have a Facebook account, Facebook is using you to get information about you and about other people you know, and that’s dangerous to the… The danger has gone to the point where they may be able to control elections or sell someone else the data to do so. It’s now beyond the point of being a threat to the individuals that are used. So you really had better not be a used of Facebook. That’s why at the beginning of every talk I give I tell people not to put photos of me into Facebook or Instagram or WhatsApp, because those are three tentacles of the same company.

One month after the just-quoted interview, Time magazine reported that the United States Congress was investigating a company called Cambridge Analytica for its role in influencing the outcome of the 2016 United States general election. Cambridge Analytica infamously pioneered a marketing technique called microtargeting. By gleaning data from prior academic research and through their own deceptive mobile app, the company was able to collate massive amounts of personal data from Facebook profiles, more data than Facebook said they ever exposed to third parties. In some cases they were even able obtain such data for people merely because they were in the friend networks of people who used the company’s app, though they had never interacted with it themselves. The key insight of microtargeting is that by gathering information about individual posts someone has interacted with on a social network like Facebook, marketers can identify target audiences with much greater precision than they can from traditional demographic data alone. Knowing what you’ve “liked” on Facebook gives any marketer much more to go on then knowing that you’re a woman, that you’re thirty-seven years old, or that you live in the suburbs of Cleveland, Ohio. What you prefer to look at on Facebook gives a much more direct window into your personality, into what you care about—and, perhaps most importantly, what you fear—than anything you would write on a government census form.

And while Facebook will continue to face public scrutiny for its role in the dissemination and microtargeting of misleading and manipulative political propaganda, and will try to use its small, overworked team of fact-checkers to deflect criticism, we know that the company is motivated, even obligated by fiduciary duty, to expand its revenues from a service that it provides free of charge to millions of people. It has even tried to expand into the banking business by creating its own cryptocurrency, Libra, which as of recording is planned for launch in 2020. As a publicly traded corporation, Facebook is legally required to not only make money but also chase opportunities to make more money for its shareholders. And if you’re using Facebook, Facebook’s advantage in the marketplace is you: your engagement, your data.

In 2019 I tried advertising this very podcast on Facebook. Facebook sent me some automated messages offering me a few dollars’ worth of free advertising credit; if I recall correctly it amounted to \$10 US. The amount of microtargeting I was able to do was astounding; I could choose to run my ad against users from specific geographic regions, who had been determined by their Facebook history to be interested in certain specific topics. I could narrow my audience by gender and age demographics. This is how advertisers on Facebook are able to reach you with an eerily specific pitch; recently my father shared an advertisement he’d seen on Facebook for a graphic tee shirt that says “Never underestimate an old man with a bicycle who was born in June.” Whoever is selling that shirt was able to use Facebook’s advertising system to market directly to men of a certain age who speak English, have an interest in cycling, and were born in June. That kind of specificity was never possible in traditional print and television advertising, and it means that advertisements no longer have to appeal to large, diverse groups of people in order to be effective. Instead of widening a campaign’s appeal to accommodate more people, Facebook encourages advertisers to narrow the target audience for each ad so it reaches the people it is most likely to persuade. It’s an advertising model that has begun to change both business and politics because it’s incredibly cost effective. That free advertising campaign I got which would have otherwise cost me \$10 netted my Facebook page thousands and thousands of views; I can’t imagine how much work it would take me to generate those figures just by making free hashtagged posts on Twitter, Tumblr, or Instagram; my interests are often a little niche, my presentation is a little shoddy, and I’m lucky if I can get a handful of likes on any given message I share that way.

So, yes, internet social media platforms are using us for our data. And though wages for Facebook have not yet become a reality, a few companies have tried to change the balance of this equation by sharing user data on an informed consent model and paying them a portion of the ad revenues. In September 2014, National Public Radio reported on three startups trying to do just that: Datacoup, Handshake, and Meeco. None of them have really made it big. The reality is that this advertising game is a competitive business, and the dominant social media platforms have two big advantages that help them maintain command of the market: one, they got here first; and two, they have the ability to make us feel as if we’re not doing business with them at all. Does the average person with a Facebook account really think of every time they log on as a business transaction? Probably not. People go on these platforms not to do business, but to catch up with friends, to speak their minds, to occupy themselves for a few moments when they don’t know what to do.

But the consequences for a company failing to maximize profits from even a wildly popular service can be far-reaching for users. GeoCities, which by 1997 had become the fifth-most popular web site in existence simply by hosting personal web pages, shut down in 2009 amid its parent company Yahoo!’s financial struggles. Because GeoCities had been so easy to use, and had offered some of its hosting services free of charge, it effectively held a massive chunk of the internet, a labyrinth of pages that largely disappeared as GeoCities went offline.

From 2012 to 2016, an app called Vine became hugely popular for its six-second video format, giving rise to a cultural phenomenon of super-short comedy, music, and journalism, inspiring dance crazes, increasing awareness of world events, and growing to over 200 million active users. In 2016, Vine went the way of GeoCities when its parent company Twitter determined the platform was not financially viable. Users of the blogging platform Tumblr who were around in December 2018 will recall the chaos that ensued when the site’s owners at Verizon, who had bought the site the previous year, unleashed automation designed to enforce its new anti-pornography policies. The bots designed to flag and remove porn and to shut down porn-focused accounts were, to use some diagnostics terminology, neither sensitive nor specific. Which is to say that pictures of sand dunes were frequently tagged as adult content while explicit photographs of dark-skinned people engaging in sexual intercourse often slipped by undetected. Meanwhile, the platform continued to have a problem with blogs openly dedicated to fascist and white supremacist propaganda metaphorically flying under the radar. The web analytics firm SimilarWeb estimated that Tumblr lost 151 million users, approximately 28% of the user base, by February. The corporate owners of social media platforms aren’t there to preserve harmonious communities. They’re there to protect their shareholders’ interests, and if that means axing a popular platform to shore up a more lucrative one, or kicking off desert landscape photographers in an imprecise effort to quickly excise pornography and the legal liabilities it presents, they’ll do it in a heartbeat.

One way that corporate social media platforms have innovated to maximize the data and advertising value they can extract from users over recent years is by liberating our feeds and dashboards from the tyranny of linear time. It used to be that when you looked at your Facebook wall or your Instagram feed you would see everything posted by the people you followed displayed precisely in reverse chronological order. The problem with this intuitive approach was that when there was a lull in posting activity, or your friends and acquaintances didn’t have anything too exciting to report, you would refresh the page, realize there was nothing new you wanted to look at, remember you had other pastimes outside the corporate social media landscape, and log off. This is a problem for social media corporations for the same reason that it’s a problem for Disney when guests at Walt Disney World in Florida leave the park to eat or sleep: when you leave their ability to make money from you leaves too.

So Facebook and Instagram and Twitter and Tumblr all used some of the latest advances in machine learning to start injecting posts they thought you might like into your feeds. And eventually Facebook and Instagram started curating your friends’ posts and showing you just the ones they thought you’d really like, and displaying them in an order that was only roughly chronological, with the stuff they thought you’d most want to see at the top. Then, if you refreshed the page, it didn’t matter if there wasn’t anything new to show you, they’d just pull out some older posts you hadn’t seen in a while, or posts that had ranked slightly lower on the hypothetical interest scale.

I feel like this curated social media experience sounds misleadingly benign without the weight of specificity behind it. Where’s the harm in YouTube or Twitter or Instagram making a few guesses about what you want to see? So I’m going to take a little detour into my personal life here and tell you a little about how this has affected my own social media experiences. I’m not about to describe a descent into the fever swamps of internet neo-Nazis and narco-traffickers, but I have been to some strange and uncomfortable places.

Back in 2010 I was a junior-year undergraduate psychology student at a small liberal arts college in Manhattan, and I was figuring out a lot of difficult stuff about myself. I didn’t yet know that I was autistic, but I was experiencing academic difficulties that would soon lead me to drop out for four-and-a-half years. A difficult breakup of a personal relationship was leading me to examine and question my sexuality in a way I hadn’t before, and the truth is that even in New York City, in 2010, it was a scary time for me.

Eventually I found that the people I related to best in that moment among a nascent asexual community on the internet, then centered around a web site called AVEN (the Asexual Visibility and Education Network), which hosted a purple-themed phpBB forum focused on asexuality, generally defined as a sexual orientation characterized by non-attraction toward other people. What I found there was a pretty varied group of people, with all different ages and interests and professions and political dispositions. But a lot of them had certain experiences in common with me, and at that moment, when I had been feeling like a disaster, like a defective person, what I found there was some hope that maybe I was actually pretty much okay overall. Many AVEN users saw themselves as existing in a social space that was naturally adjacent or connected to the broader LGBTQ community, and this was the contingent that I related to most. In fact, there were so many trans and non-binary users on the forum that their subforum eventually grew into a separate forum all its own.

I was finding that the internet was allowing me to connect to people with shared interests and experiences that I had trouble finding offline, and I wanted to expand my internet social circle beyond AVEN. A lot of twenty-something-year-old AVEN users like myself were starting to use Tumblr, and its freeform style, which allowed images and long text posts which could be easily copied between users, allowing for a mix of soapboxing and curating images to match a personal æsthetic, appealed to my sensibilities. So I started a Tumblr blog. I don’t want to be too dramatic about this, but Tumblr did eventually lead me to people with whom I formed long-lasting real-life relationships that have defined the shape my life is in today. It’s where I developed a lot of the interests I still have, in sociology and various kinds of art. It also eventually became the bane of my existence and I haven’t properly used it in years.

When I arrived on Tumblr the primary method for discovering other people’s blogs was through tags. At the time, opening a tag would default to showing all the posts that had that tag in reverse-chronological order, no matter how few people had been engaging with a particular post. Unlike, say, Reddit, it didn’t have any upvote/downvote system or community moderation; whatever someone wanted to post with that tag would be visible to users viewing that tag unless it ran afoul of the site’s fairly narrow terms of service. Organizing things by tag like this was a fairly successful way to organize people around fandoms and interests. For me this meant I could find other people I related to through this broad internet community of asexual people while also engaging with people on other shared interests, from sociology and world history to film and photography.

That tag system brought people together quickly but also encouraged bizarre social skirmishes. Early on I found that the tags used by Tumblr’s nascent asexual community were also frequently used by a small but very vocal group of people who were there to argue vehemently that the terminology and symbols used by this community, and really everything about how the community defined itself, was a dangerous and harmful encroachment upon the broader LGBTQ community, making meme pages with names like “Homophobic Aces,” explaining at length their belief that the grayscale-spectrum inverted triangle used in AVEN’s logo was maliciously appropriated from lesbian culture, and voicing their concerns that the asexual community would tamper with comprehensive sex education. In real life I was drawing emotional strength from a newfound feeling of connection to the broader LGBTQ movement, but in these little Tumblr tag virtual spaces I was daily facing accusations that I was out to derail or destroy that movement.

It would probably have been best if I could have just avoided this micro culture war altogether, but at that point in my life, I was hardly capable of socializing at all outside the internet, and not only did the largely unmoderated Wild West of Tumblr tag communities do little to help anyone disengage from these conflicts; over time the platform tended to enable or exacerbate them. The primary method for people to engage with even the posts they really disliked was to “reblog” them, to copy them onto their own blogs, with the option of adding their own commentary at the end of the post. So if someone whose blog I followed wanted to publicly object to a post they had seen, the post would show up on my Tumblr dashboard. This poses the same problem as Twitter’s “quote tweet” mechanism: at best, it tends to put posts directly in front of the people most likely to find them upsetting; at worst, it encourages widespread performative outrage that poisons the whole culture. And then, to make matters worse, Tumblr would start showing me the most incendiary, unnecessary posts, stuff I had no interested in, just because it was posted in a tag I had followed and had been reblogged a lot. I’d stop viewing the tags and stick only to the blogs I was already following, but these posts began to be inserted at random between others on my dashboard, marked with something like “recommended for you.”

Over time it became too much for me to casually engage with Tumblr at all. The awkward little skirmishes around the asexual community turned into a whole Tumblr subculture of their own, dubbed “ace discourse,” and though I carefully curated the blogs I followed, by 2015 the “ace discourse” had metastasized seemingly at random to other parts of Tumblr and would show up on my dashboard because the person running a blog I followed about architecture or socialism or something would receive an anonymous message asking for their stance on whether asexual people belong in the LGBT community. And it got weirder and weirder overall. 14-year olds would take up ace discourse as a hobby and create secondary blogs just to indulge in it. A few people were exposed for creating false personas. Threats of violence were exchanged and bloggers dared each other to commit suicide.

And these unhealthy dynamics increasingly played out in other conflicts on Tumblr too, even as the demographics of Tumblr seemed at least to me to be trending younger and younger. I’d log in to check up on my boring adult interests and find myself awash in a sea of fandom bloggers in their early teens accusing each other of supporting pedophilia over their disagreements about which characters in a TV cartoon should date each other. Gradually, I decided I was officially too old for this shit.

In 2016 I tried out Twitter again. I had used it a little, back around 2010 or so, when the character limit was 120 and I sent and received tweets by text message, using it to post odd haiku or follow writers I thought were interesting, but in 2016 I recognized that the long paragraphs of discussion and argument from Tumblr encouraged addictive behavior from me because they played into my fear of leaving things unfinished; I couldn’t just stop reading in the middle of the post. And my first child was born in 2016, so that wasn’t going to do. At first I just drastically cut my social media usage altogether, because it was hard to use with the newborn around anyway. But eventually I tried replacing my Tumblr habit with an Twitter habit. At first it seemed to be an improvement. It was easier to break away from Twitter simply because the posts were all short, limited to 140 and later 280 characters. But 2016 turned out to be a bad year to be an American on Twitter.

In the lead-up to the 2016 U.S. presidential election, the norms of respectability and acceptability in American politics completely imploded. Open fascism and white nationalism were creeping out of the fringe and into the mainstream. And, as I explained before, the quote-tweeting dynamic of Twitter ensured that some of the liberal or left-leaning folks I followed there kept putting the worst of this trend all over my Twitter feed so they could respond with witty rejoinders. And it only got worse when the election was over. I had largely escaped the ace discourse, but now I couldn’t log on to Twitter without seeing the endless stream-of-consciousness ramblings of the American president-elect interspersed with recriminations over who was to blame for his new-found prominence: the ungrateful “Bernie Bros”, an incompetent Hillary Clinton, uneducated bumpkins of a backwards rural heartland… It was a sort of grotesque political soap opera acted out by empty caricatures, largely divorced from the reality of anybody’s daily life.

Twitter didn’t do much to moderate this scene and encourage good-faith discussion. Twitter doesn’t have any business interest in good-faith discussion. To a business like Twitter there is only engagement, advertising revenues, and shareholder satisfaction.

It wasn’t always this way. Social media on networked computers was not always a business, let alone big business. In 1978, computer hobbyists Ward Christensen and Randy Seuss launched the Computerized Bulletin Board system. Essentially this was a specially configured computer connected to a telephone line in Seuss’s Chicago home; someone who wanted to connect to the bulletin board system or BBS would dial its phone number and place their phone in a little cradle on a device connected to their computer; [phone modem sound] this allowed them to share messages with the BBS, to make posts where anyone who connected to the BBS would see them and to peruse what other users had written. BBSes became quite popular in the 1980s, particularly in places with a high concentration of computer users in the same local telephone-calling area, because at the time users in places like the United States could rack up costly telephone bills by placing calls outside their local area. Like many mainstays of computing culture in that era, they were run primarily by and for hobbyists. There are some old-school BBSes still running today, but if I had to draw an analogy to something that remains popular now, I’d say a BBS is a little like Craigslist: visually plain, mostly textual, and with a focus on recent and geographically proximal content. It was an attempt to virtualize a literal bulletin board, a public place where passers-by can leave a note for their neighbors.

In time the mania for role-playing brought about by the success of Dungeons & Dragons collided with the rise of networked computing to create a more specialized kind of social network. These were often known by acronyms like MUSH, MUCK, MOO, but probably the best-known term is MUD, for multi-user dungeon. Like the early BBSes, MUDs typically operate over a simple text-based interface, but instead of a simple bulletin board interaction they provide an experience more akin to a multi-user version of popular 1980s text-based computer games like Zork, with the added fun of players having the ability to create new objects and places within the game’s world as they play. But, like BBSes, these are hobbyist spaces. The typical MUD is no more a corporate enterprise than the typical group of local D&D role-players.

Even network computer systems that were never intended as explicitly social spaces would come to foster social communities. Beginning in the ’70s, universities that offered the best available computing resources to their students and faculty adopted shared Unix systems. Unix is an operating system, an ancestor of Mac OS, Android, and the Linux systems in wide usage today, that was originally created for the Digital Equipment Corporation’s PDP-11 computers, popular amongst academic institutions, and later rewritten to work on a wide range of computers from competing manufacturers. One thing Unix did very well that made it so popular was facilitating multiple users to performing their work at the same time. The best computers were expensive, and the ability for multiple users to run commands on them simultaneously was crucial to the research of the time. With the rise of Unix, institutions could ditch the scheduling systems of old, where students would have allotted time slots during which to run and debug their programs. Over time, computing hardware systems in academic institutions would become more distributed across campus, with consoles in disparate locations connecting to the same shared servers.

A side effect of these multi-user systems was that people in these academic institutions began to meet each other in a brand new way: virtually, as fellow users of the shared Unix system. The standard Unix command who provides a list of currently connected users and the finger command lists details about an individual user, like whether they have any unread email, and the contents of a special “plan” file that contains whatever the user has decided to put in it, like a precursor of the Facebook status update. And while users could make all their files private, accessible only by themselves and the system administrators, often they didn’t, allowing others to poke around in their “home” folders and read some of their files. As these academic institutions began to connect to their internet, students and faculty alike might have the opportunity to use a special folder within the home folder to host their own pages on a university web site. Email and eventually IRC allowed users to reach out to each other directly without having to ever see each other in-person. And while their nominal purpose in using the school computers was educational, students could hardly be expected to keep everything strictly business. Zork, which I mentioned before, started as an experimental game called Dungeon created and hosted right on the Massachusetts Institute of Technology’s computer system; its players were other students at MIT. Plan files contained ASCII art of cows. Simply typing the word fortune into a terminal would bring up a pithy quote from Ambrose Bierce or Dave Barry. So, yes, fun was allowed, and actual friendships could form. I’ve even worked with someone who met his wife through their large university’s Unix system, before they got to know each other in person.

In 2014, a writer and techie named Paul Ford, inspired by Stevie Nicks, decided to re-create this social Unix phenomenon for fun. See, Paul Ford was reading a 2004 blog post by the singer-songwriter and former member of Fleetwood Mac and noticed that she used the tilde a lot. That’s the little wavy line that appears over the letter ñ in written Spanish. So, as he explains in his essay “I had a couple drinks and woke up with 1,000 nerds”, he turned to his wife Mo for an explanation:

“That’s a west coast thing,” said Mo, who grew up in California. “You use tildes instead of colons or dashes, it’s more like handwriting.”

I shared this fact on Twitter. A website-managing friend of mine promptly wrote back: “tildes are only ever properly used in front of usernames on shared hosting.” Wait, what?

So! Back in ye olden Inter-Net tymes of the 1990s, if you were a (then youngish) nerd like me, you’d get an account on some server called CyberFox.net and your web address would be http://CyberFox.net/~vixen. (You were “vixen.”) And you could put some web pages at that address.

The early personal web grew up around these little “tilde sites”; that’s what preceded blogging. The “~” is a little like the “@” on Twitter—a shortcut that says: “Here is a person.”

By now I was well in the bag and listening to Stevie Nicks (sounds like she’s singing, hoo [baby], hoo [baby], hoo), so I tweeted out, “YES also i (sic.) just registered http://tilde.club and will given (sic.) anyone a shell account who wants one.”

A “shell account” gives you access to use a computer. When you log in a “shell” program runs that lets you issue commands. This is what hackers type into in movies. From the humble command line you can write things and save web pages.

So Ford had rented a relatively cheap virtual private server—basically just some space on a shared computer somewhere in Virginia, which can be obtained on the internet in a matter of minutes. The following morning Ford woke up to 100 account requests and got to work letting people in. But it’s what they did with their accounts that made the whole experiment interesting. From Ford’s essay again:

People logged in and started live-chatting all-text pictures of dragons to everyone on the computer. Then they switched into their text editors and started to make web pages, right there on the server, just like I’d hoped.

Within the first week of tilde.club’s existence, its users had created a slew of web pages, blogs, code that could show web visitors which users were logged into their shell accounts, code that could show web visitors which users were using too much disk space, a geographical map of user locations, an infinitely recursive animated cartoon talking cow, a spooky interactive picture of the musician Enya where the pixels that make up her face gradually swarm into formation out of a high-entropy sort of cloud and you can scatter the pixels with your mouse as if you were disturbing ripples on a pond with your hand, a tool just for playing four random rock songs from YouTube at the same time, and an animation of the Abominable Snow Monster from the 1991 Microsoft game SkiFree eating other tilde.club pages. What had been for Ford a strange late-night impulse purchase became a sort of community overnight. People volunteered their time and money to help manage the unexpected flood of users, and by now there were a number of copycat projects, community servers much like tilde.club loosely connected in something called the tildeverse. I personally have an account on a tilde server called tilde.town.

It had been only about two weeks since tilde.club started when Ford wrote his essay, and already he was being asked about its money-making potential:

In the last week many people have asked me: Is Tilde.Club a social network? Is it a company? Is this a product? A minimum viable product? What did you do to build it? People have asked to talk with me about it, to interview me about it, and asked for my plans and goals. The site has been listed on Product Hunt, which is a website that itself just received \$6.1 million in venture funding, so that it can fully carry out its mission of making a list of other websites that receive venture funding. And to that all I can say is—

Tilde.club is one cheap, unmodified Unix computer on the Internet.

That’s it. That’s all it is. It is no more than that.

The tildeverse is rife with old-fashioned software and memories of earlier of computing, but it is not itself an exercise in nostalgia. It is moreso an experiment in truly social computing: what if we were to simply share computers the old way, the way they did in the ’70s and ’80s, not necessarily because we miss those decades of computing but because their methods did not rely on corporate mediation of our social spaces? What can those old models of social computing accomplish when they are connected to the full power of today’s internet? It’s a sort of “stone soup” way of doing things. The Unix server at the core of tilde.club was created with only the same old tools that are standard on any old computer system these days. There was nothing about it that made it uniquely suited to be the catalyst for a new internet community, nothing that Facebook, Twitter, or Google could acquire and monetize. What actually made this a community was the participation—the ideas, art, and code—of all the users who showed up, lured by the promise of an anarchic do-it-yourself shared computer where nothing they did would need a business justification.

Now, you can probably tell what my feelings about the tildeverse are. I think it’s pretty cool. But I have to acknowledge it’s not for everybody. Specifically, as welcoming as a tilde server can be to amateurs and newbies, no one would be there if they weren’t at least willing to learn how to do a few things in a Unix command-line environment. And I’ve been told there are a number of people in this world who aren’t interested in using a Unix command line. I know it sounds strange, but trust me. These people exist. Is there a non-corporate model of social computing that could work for some of these people? Yes! I’m so glad I asked.

Since the early day’s of Twitter’s popularity after it launched in 2006, a lot of hobbyists and volunteers have tried to produce various kinds of Twitter-like or Facebook-like microblogging platforms that ran in less corporate, more community-oriented ways. After all, storing and displaying user posts on a web site isn’t the most technically complicated thing to do, if you ignore for a moment the challenges of managing millions of users on a single site. For a while this scene was a little chaotic, but over time these different platforms became better at sharing the things people posted with each other, so by about 2010 it began to matter less and less that all of the sites that hosted these experimental microblogging accounts were small community-funded projects that couldn’t handle the vast number of users Twitter was accruing; you could post a status update on your own little GNU Social web site about the new bubble tea café around the corner and your friends who were following you from their Friendica and Diaspora web sites could read it and reply. What was emerging was a so-called “fediverse”, a network of various sites and apps and services that might look different, might be owned and operated by lots of different people for lots of different reasons, but are nonetheless “federated,” meaning that what a user posts on one of them can easily propagate to their followers on other sites & apps in the network, much as you can use your riseup.net email address to email someone who uses GMail, or still has a HotMail address, or has some kind of corporate email account.

But use of these platforms remained somewhat fringe for a few years. Things really started to change in 2016. It was in March of that year that a new contender in the space of these free software Twitter-like microblogging platforms was first released, a project led by developer Eugen Rochko called Mastodon. It’s hard to pin down exactly why Mastodon took off in a way that platforms like GNU Social hadn’t, but there are a few things that stick out: Mastodon started out with a pretty slick user interface, one that would seem to belong on a well-funded corporate platform if it weren’t for a complete lack of advertisements, strictly reverse-chronological timelines, and a content warning feature that allows users to hide posts about sensitive topics behind a “read more” button. From an early stage, it emphasized easy moderation tools so that people running individual Mastodon sites could form moderator teams to enforce their specific rules. The different Mastodon sites were fairly good at synchronizing posts and user profiles with each other, and quickly got better at it. And the whole user experience was just familiar enough that some people who wandered over from Twitter got the hang of it and stayed, even though mechanics, ethos, and prevailing mores were fundamentally different. But I also have to wonder if the relative success of Mastodon might be partly because it happened to come at just the right time, when the demand for alternatives to Twitter and Facebook was reaching new heights.

I was not the only person who was very tired of Twitter in 2016. While Twitter dragged its feet on the issue of growing, open white nationalism, I and at least a few other people migrated over to the fediverse in search better-moderated communities. I can’t speak for everybody who did this, but for me it basically worked out. I found a niche for myself with people who were capable of thoughtfully engaging with politics without wallowing all the time in this spectacle of metaphorical mudslinging that was all over my Twitter feed. I also found myself more easily meeting and following people from outside the U.S. who brought to the conversation an understanding of our international common causes. So for me, the whole community moderation model was working. Eventually, it would be put to the test.

In October 2018, Robert Bowers, a frequent user of a Twitter-like platform called Gab, killed eleven congregants and wounded six others in a synagogue in Pittsburgh, Pennsylvania after posting his plans on Gab. Gab catered heavily to former Twitter users whose white nationalist propaganda was brazenly violent enough that it eventually got them banned. In the wake of this mass murder, Gab was taken offline, and though its operators were determined to restore Gab, they struggled to maintain hosting under the deserved stigma created by this mass murder. It wasn’t long before they determined that the decentralized model of the fediverse could lend them the resiliency they were looking for. Gab was reinvented as a Mastodon community, first with one Mastodon server of its own, then more, so they would no longer be vulnerable to getting shut down at a single point of failure.

You might expect, at this point, that my Mastodon timeline suddenly became flooded with these Gab Nazis. But actually, I didn’t hear a peep from them. This is because the volunteer administrators of the Mastodon server I was using quickly joined with others to identify and block Gab servers as soon as they appeared as a strategy to isolate Gab from the mainstream fediverse. It’s a strategy that has proven successful; while we don’t have the power to completely wipe Gab off the internet, thanks to these community efforts Gab is effectively cut off from the social media spaces I use, sandboxed in their own fractured, insular network.

This is not to say that the fediverse has cured all the ills of the social internet. Addictive behavior, performative outrage, and unproductive drama between moderators still happen. The relative anonymity of these social spaces still disinhibits some harmful speech. But what the fediverse doesn’t have is business-driven features designed to drive engagement of any kind for the sake of advertising revenue, features that actively encourage these harmful behaviors. And it’s nice!

I took a very long time to write the script for this episode. A lot of stuff has happened in my life while I’ve been working on it intermittently: I finished my bachelor’s degree, made some plans and changes with my family at home, and briefly visited San Francisco. Also a pandemic emerged, so there’s that. The physical isolation many have endured in the interest of public health means that a lot of us are relying on technology-mediated socialization more than ever. While people have been isolated in their homes, trying to make sense of their newly disrupted lives, a lot of them have encountered misinformation about what causes the disease, how it can be prevented and treated, and how dangerous it is, statistically speaking. Despite the slow and haphazard efforts of corporate platforms like YouTube to remove some of the worst offenders, well, I know my uncle has seen Plandemic, the widely-circulated false documentary encouraging lots of people to act counter to public health advice, and has circulated misinformation about debunked cures for COVID-19. I know this because he shared it on his Facebook timeline, and if I truly believed any of this misinformation after he had exposed me to it, I probably would have shared it too. Disturbingly, people who monitor the QAnon movement, a quasi-religious, largely far-right cultural phenomenon driven by conspiracy theories about widespread pedophilia and cannibalism perpetrated by American public figures, report what looks like a reinvigoration of QAnon fed by COVID-19 misinformation.

Occasionally, in order to combat misinformation or hate speech giving the platform a bad name, Twitter will ban a bunch of accounts, or Facebook will delete some groups, but any action like this is a weak attempt to overcome an intentional design feature of these platforms; they’re all designed to retain our attention by showing us the things that emotionally provoke us in some way. These platforms’ failure to contain misinformation has real-world consequences. Disorganized movements like QAnon that probably could not have grown without having their propaganda amplified on Twitter, YouTube, and Facebook have encouraged people to challenge public health policies and in some cases even kidnap their children involved in custody disputes.

I’ve tried to make the case here that non-corporate alternatives to the dominant internet social media platforms exist, that they can be used to improve the quality of our social behaviors online, and that some of them, at least from a technical or user interface perspective, are ready to replace those dominant platforms for day-to-day usage. But even if these platforms solve all the technical issues and give us flawless interfaces, there will be one big hurdle to overcome: the network effect. The network effect is why so many people will tell you they’re tired of using Facebook but still have active Facebook accounts, or say the same of Twitter. Their social networks, in the sociological sense—that is, the complex web of people to whom they are socially connected through work or hobbies or family or geographical proximity—are all mostly on Facebook or Twitter already. So if you’re looking for something to do about improving the internet social media landscape, my advice to you is to be an early adopter. Try one of these newfangled platforms like Mastodon or PixelFed and see if you can connect to people of similar interests there. Maybe someday, when your old friends decide they’re fed up with Facebook and Instagram, they’ll be able to find you there, on this other thing. And remember, these are community-driven projects. Maybe you’ll have feedback that can improve them by the time your old friends arrive.

This episode of The World on a Wire Show was written and narrated by me, Dominique Cyprès. The opening theme music was “Come Inside” by Zep Hurme featuring Snowflake. The closing theme is “Start Again” by Alex Beroza. Incidental music used in this episode includes “reNovation” and “reCreation” by airtone featuring mwic. The modem sound effect was recorded by guitarguy1985 on freesound.org. Get the latest updates on The World on a Wire Show at patreon.com/lunasspecto; that’s Lima Uniform November Alpha Sierra Sierra Papa Echo Charlie Tango Oscar. You can also reach me by emailing lunasspecto@gmail.com.

This work is licensed under a Creative Commons Attribution 4.0 International License.

Hey! This episode touched on some really big topics that could be impacting your life right now for all I know, so I’d like to throw a few extra resources your way if you’re interested. If you’d like to explore the topics of emotional self harm, hate speech and incitement to violence on the internet, I recommend Natalie Wynn’s August 2018 YouTube video “Incels” for her channel ContraPoints. If you’d like to explore how performative outrage—and outrage in general—functions in a capitalist market, I suggest Peter Coffin’s June 2019 video “The Outrage” on YouTube. And if you would like to know more about the particular mess that is QAnon right now, I suggest Adrienne LaFrance’s essay “The Prophecies of Q” published in the June 2020 issue of The Atlantic. Or, if you’re looking for a whole podcast on the topic, there’s The QAnon Anonymous Podcast.