- A brief respite from editing posts on polemic books to bring you news about the ex-bird site – news about its ex-users, that is.
- The migration from X (formerly known as Twitter) to Mastodon gave researchers a golden opportunity to record data on the collective action problem. Spoilers: it’s pretty good news.
Yesterday, The Verge posted a special report on the year Twitter died:
https://www.theverge.com/c/23972308/twitter-x-death-tweets-history-elon-musk
Up until the point that Elon Musk bought Twitter, I’d tell you that if I never saw Twitter again it would be too soon. At the same time, I continuously reflect on a conversation I had during the apex of the first lockdown with a friend of mine studying politics. Love it or hate it, he said Twitter was the place to go for journalists – professional or otherwise – to break stories and search for the next scoop. He said, as I recall, that no matter how bad things get, it would never not be valuable for that.
He was right – and then Elon Musk bought it so he could speedrun the enshittification cycle in 0.5x A presses. I, as ever, refer to Cory Doctorow’s auxiliary theory1 to explain how platforms play both end-users and business customers to sustain themselves:
https://pluralistic.net/2023/01/21/potemkin-ai/#hey-guys
The call to action has hit the mainstream. A recent video by Kurzgesagt (In a Nutshell) raised awareness about platforms going to shit, and how we seem to become more divided by the day. At time of writing, it has over 5.6 million views:
There can be no doubt that, across the board, people know something’s not right. The blame seems to be shifted to shaky addiction metaphors caused by the ａｌｇｏｒｉｔｈｍ™. Don’t get me wrong, a platform’s sorting algorithm does something. But that something I don’t think is significant enough to be treated as some omniscient force that either brainwashes you or knows exactly how you’re feeling to keep you hooked. Let’s not give the platforms – or their business customers – too much credit here.
If someone says, “God, I hate this platform, but I can’t/won’t leave,” they are entitled to information and resources to at least understand why. On this point, Kurzgesagt’s video is, in the nicest of ways, a bit shit. The thing about filter bubbles is mostly correct, and the call to action is absolutely warranted, but on the whole it’s really a bunch of observations that aren’t new and don’t offer any tangible solutions.2 The whole thing reeks of, “The platforms are dying – that’s sad, because there are no alternatives.” This video only furthers the beliefs that people already have, and does little to challenge them.
The mentioning of the old, good(ish) internet, and how it was a bit like local communities, was the perfect segue into bringing up last year’s mass Twitter migration to Mastodon (and other ActivityPub-powered platforms). It’s a staggering example of a collective behavioural change at large that individuals and groups alike have participated in voluntarily. Studying this sort of thing in the wild is tricky business, so it’s even more staggering that it’s on record:
https://fediversereport.com/study-on-the-twitter-migration/
This post made its way into my Mastodon feed earlier today.3 It’s a TL;DR of a recent study on the drivers of social influence in last year’s migration. Researchers Lucio La Cava, Luca María Aíello and Andrea Tagarellí collected data from 75,000 Mastodon users, and used a COVID-era epidemic model to investigate the extent to which social influences – like publicly posting your exit from Twitter – had an effect on other users leaving Twitter.
For all intents and purposes, social media silos act as intermediaries – from user to user, and or business customer to user. It is a social tool not unlike the telephone in that its value grows in proportion to the number of people using it. A platform with ten registered users means you can interact with ten users. Not a lot of people, and not a lot of reason to use it. But a platform with over three billion active users? If you’re not on there, you’re missing out. Economists would describe platforms as having network effects:
https://www.investopedia.com/terms/n/network-effect.asp
I know the power of network effects first-hand through my Dad. He’s a self-employed painter and decorator of 35 years’ experience, and like many self-employed tradespeople, he runs his small business on Facebook. That’s where the customers are, and for the customers, that’s where he is. While he is asked for a quote by many a customer through text, phone call, or email, if Facebook did nothing to increase his reach, he wouldn’t be on it.4 Users make the service – the service doesn’t make its users.
Say that Facebook was to shit the bed tomorrow, completely out of the blue. Say that Meta pays Apple and Google under the table to enshittify the Facebook experience for tradespeople anywhere they might access Facebook from – mobile apps, Safari, Chromium-based browsers, etc.. They introduce a paid verification system, and make little twiddles here and there to self-sabotage their app’s user experience for non-verified tradespeople. A post-viewing limit here, a bit of unresponsiveness there.
Anyone who doesn’t cough up a monthly subscription fee to become a trustworthy, ‘Metafied’5 trader, and rent space for their business, will be SOL. They’ll find it harder and harder to do business with their customers, until they either pay for verification, or take their business elsewhere.
Facebook, in this example, would be abusing its small business users to make a bit of coin. But it’d still dominate as the platform where a small business would find its largest available reach. If Dad found this unfair, and refused to cough up the fee (which, given the precarity of his job, would be no small sum), he’d face a two-pronged problem: Persuade other tradespeople to move to a different platform that treats them better, and persuade his customers to do the same. Sociologists would categorise this as a collective action problem:
https://en.wikipedia.org/wiki/Collective_action
As for the consequences, economists would describe them as switching costs:
https://www.investopedia.com/terms/s/switchingcosts.asp
Network effects are how Big Tech platforms got big. They remained big by making it as difficult as possible to leave and take your connections with you (among other factors).
Something moved us to migrate to Mastodon. Until now, all we’ve had to go on was stories, and while qualitative user experiences are valuable, they needed to be paired with quantitative statistical data to get the full picture. The researchers above have done this, and the results are very interesting.
Their data set specifically contained 2 million tweets – 1.3 million of which are in English – from 500,000 accounts clearly declaring their exit from Twitter, their migration to Mastodon, and links to those new accounts (2.5K of which declared a Twitter handle in their metadata). The timeline for this set spans from the 26th October 2022 to 19th January 2023.
If the collective action problem holds true, then using Twitter to declare your migration has a social influence on your followers, thus influencing some of them to join you. Posts about ditching Twitter beget more posts about ditching Twitter. The SIRS model, used to simulate disease spread, can also be used to map to other kinds of data – like Twitter posts – “to simulate information diffusion within social systems.” To use an old and crusty term from the 2000s, the researchers measured how how quitting Twitter went viral. It’s not unlike tracking hashtag usage over time, or the spread of conspiracy theories, as the authors cite in their sources:
https://arxiv.org/abs/1603.00074
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0256179
Based on these tweets, the researchers found that the spread of Mastodon account creation was highly infectious. The cumulative number of users who signed up over the first ninety days of the Musk takeover and stayed on Mastodon increasingly outnumbered the number of users who threatened to switch and stayed, as well as users who switched to Mastodon and came back to Twitter.
Not all of the communities recorded were equally infectious. In this context, ‘communities’ refer to groups found using the Louvain algorithm (we might say we’re a part of ‘Art Twitter,’ ‘Kirby Twitter,’ or ‘Smash Twitter,’ but network scientists have figured out how to express this statistically). Those communities who migrated faster than the others were found to have some things in common:
I. The more users there are in a community, and the more connections they have, the more difficult it is to reach the incentive to migrate. The opposite is also true – less friends means less people required to migrate to Mastodon to meet the incentive.
In my post reflecting on my exit from Twitter, I mention a friend of mine who, like many teens growing up in the late aughts/early 2010s, used Skype to talk to his friends:
https://brologue.net/2023/11/15/i-want-my-twitter-friends-back/
Skype was horrendous, but they’d used it for years and didn’t want to switch to an alternative. Then, Discord came along in 2015 with an ad campaign eerily mirroring Facebook in its early years punching up against Rupert Murdoch’s MySpace: Join Discord, where we’re not Skype, or Teamspeak, and our group chats are E2E encrypted:
https://web.archive.org/web/20151225064513/http://discordapp.com/
This is a prime example of a tight-knit community where the switching costs are high. Either everyone switches, or no-one switches – a decision clearly triggered through strong social influences.
II. Reiterated commitment: Members in the community continuously repeated their plans to migrate. The more this intention was repeated amongst community members, the more users took up the call themselves.
To capture the Tweets used in the dataset, the researchers created a complex search query and submitted it to Twitter’s API. Here it is in full:
@joinmastodon OR mastodon OR #TwitterMigration OR #RIPTwitter OR #TwitterTakeover OR #TwitterShutdown OR #TwitterIsDead OR #MastodonSocial OR #LeavingTwitter OR #ElonIsDestroyingTwitter OR #MastodonMigration OR #Fediverse OR #Mastodon OR #TwitterAlternative OR #DecentralizedSocialMediaTwitter’s always been known as a powder keg for blowing up social phenomena. The hashtag’s usage today is a kind of détournement from their use in the blogging days. What once was used by the individual author to organise their posts was adapted as a signal to other users – “I’m posting about this, and if you find it interesting, you should post about it too.”
III. Of those communities where social influence triggered rapid migration, people were more likely to co-operate when social discourse reflected a sense of shared identity, and “[emphasiszed]… exchanges of factual knowledge.”
Back when I was hip and played Smash, I used to be a moderator in a quasi-official Discord server6 for Olimar players. Some time after that, I stepped down from the mod position and opened the PikuoriParadise Twitter account. It quickly became a safe haven for me and others on Smash Twitter to talk about how horrendously bugged Olimar was. I can’t say a thing for the other character Discords, but the Olimar server, to me, always had that shared sense of identity. We played the same character, shared our troubles, our triumphs, and exchanged esoteric technical knowledge not applicable to any other character.
When someone in the States discovered Nintendo’s customer service was actually listening to feedback about Ultimate’s bugs and glitches, I decided to spearhead a campaign to get Olimar his dues. This wasn’t without precedent, mind you – Olimar was arguably an even more buggy mess in Smash 4, and top players kicked up a fuss to no avail. I’m not sure how influential the #FixinthePikmin campaign was, but as it ran its course, we saw some of Olimar’s most egregious oversights addressed, including a bug with his Up Special that’d been carried over from Smash 4. Fucking carried over!
https://twitter.com/pikuoriparadise/status/1340023577983201280
My Twitter was but a tiny haven, and most, if not all of the people who followed me were more densely connected to the rest of the Smash community. The switching costs for me were high, but for my followers, they were even higher, so they didn’t come with me to Mastodon.
https://geekdom.social/@bigolifacks/109585532618447639
Grassroots is grassroots – whether that’s campaigning to fix a bug in your favourite game, or moving to another platform, the researchers note the combined influence of identity and knowledge as having a greater influence on behavioural change than community density and commitment.
The bottom line is this: When it comes to anything that doesn’t get immediately branded as ‘cancel culture,’ Twitter activism has always been accused of being all talk, no action. What this study shows, however, is that there is credence to the idea that not only did an external pressure move enough people to take action, but that this call to action grew organically through discussions about what the alternatives are; what the switching costs would be; and clear signposts to signal user departure. Twitter users can and do take action, with tangible results. If this study can be replicated by other researchers, then it’s a strong argument in favour of using federated social networks to enable mass co-operation.
If you see someone write, “God, I hate this platform, but I can’t/won’t leave,” this study is the sign you want to tap. Twitter’s shit for the birds – let’s leave.
Mark Zuckerberg’s a fucking chancer, by the way. As I was writing this, The Verge put out another story – apparently, Threads is federating:
https://www.theverge.com/2023/12/13/24000120/threads-meta-activitypub-test-mastodon
I can’t believe it. I just can’t believe it. The Zuck has come down from Mount Olympus to bestow upon us the gift of interoperability because he heard it was a good thing, actually–
It doesn’t sound like you’ll be able to post from Mastodon to Threads, for instance, and you can’t move your account between services.
Ceci n’est pas de l’interopérabilité. 🚬
- Not that enshittification needs any such validation, but it should’ve been the word of the year. If you need a SFW version you can say to your boss, your teacher, your mum, or in Christian gaming servers, say “platform decay.” ↩︎
- I don’t know about you, but I swear I’ve seen the ‘How to Destroy the Universe’ title used on like five different videos before they change it and the thumbnail to something else. When it comes to videos on cataclysmic world-shattering phenomena, Kurzgesagt milk it hard. ↩︎
- The site looks like a blog, but it’s actually running on ActivityPub. That’s why I could see it in my Mastodon feed as a Mastodon post – federated platforms can interact with each other, and end users can engage with content as they like it! ↩︎
- Assuming reels did not exist. Dad likes a reel. Or fifty. ↩︎
- See what I did there? ‘Metafied?’ Meta + verified. ↩︎
- Official in that it’s affiliated with SmashBoards – not official in that it is in no way affiliated with Nintendo.
↩︎
Wash That X Site Outta Our Hair is licensed under a Creative Commons Attribution-ShareAlike 4.0 International license.