So long, and thanks for all the birdseed

I hadn’t really planned to write this today, but it seems like — if it’s going to be visible at all via the main place I’d planned to link it — I really needed to get a wriggle on. Obviously I’m not trusting the actual content to that platform, because…

Well, put it this way: My job frequently boils down to monitoring and analysing trends. Not fashion trends (god, would I ever be bad at that!), but rather trends developing in complex systems. Trends in weather (often, my best decisions come not from interpreting a weather forecast as published, but by looking at a series of weather forecasts and observing the trend in that series of weather forecasts), trends in a flight path (am I getting closer to or further away from the ideal place for me to be at this particular moment), and so forth.

Having identified a trend that is of significance, you need to make a call about what (if anything) you can or should do in response to it. Obviously if my flight path is trending away from the ideal, then that warrants changing something to reverse the trend. If weather conditions are trending poorer, then I need to start making plans about where we might go (other than our intended destination), as well as identifying in advance at what point each decision will need to be made.

There is no need for a great deal of expert analysis to assess what is happening with Twitter, nor to make a reasonably sound prediction about where this ends. In short, it’s dead. Even if you could wave a magic wand and reverse all the short-term effects of a certain emerald-moneyed billionaire getting his hands on it, it would still be dead. It’s probably been dead for some time — we’ve just been flogging that horse hard enough to convince ourselves it was still moving.

I’m not going to delve too much into what killed Twitter, although it is worth noting that a lot of the elements that killed Twitter have the potential to do exactly the same to any number of other social media platforms: any algorithmically-driven social media is going to suffer from the same problems that news media does in representing the world around it, no matter who the players are. Bad news drives clicks and views, so news media is almost all bad news. Anger and hatred drive engagement, so algorithms designed to maximise engagement will promote hatred and anger.

You can’t leave the visibility of content up to such an algorithm without causing harm to the human participants. And slapping certain policies on top of an algorithm like that can’t fix it, even if you have a well-resourced and dedicated team in charge of enforcing those policies. You don’t fix nazism and white supremacy by applying a blanket policy of “no threats of violence or hate speech”, because then well-behaved, calmly spoken nazism and white supremacy is perfectly acceptable… and as far as the algorithm is concerned, worthy of promotion, because look at all the engagement it generates!

So, like it or not, Twitter is dead (or at least, mostly dead… and Miracle Max isn’t coming to its rescue). And that’s a shame, because it’s done a lot of good things in its time. There was a certain democracy of access in being able to tweet at a celebrity or politician or business and have them potentially see it and respond. I launched and promoted a political campaign via Twitter (not seeking election myself, I hasten to add) that, while largely symbolic, still had a measurable effect on the votes in the election in question. Hopefully that effect, in turn, communicated something of the unpopularity of a policy being championed by the target of that campaign. And that’s something I couldn’t have done without Twitter.

I’m reading plenty of stories from people talking about the things they’ve got out of Twitter over the years: connections made that would never have been made otherwise, relationships and families that owe their very existence to such connections, movements launched and substantial positive impacts on people’s lives and livelihoods. I think it’s important to recognise that Twitter really has enabled those things over the years. In its earlier years, Twitter represented a lot of the best of what social media had the potential to be.

But as it gained in influence (and therefore potential profit for those who could control and exploit it), so the very things that made it special declined. With money to be made, everything became about chasing and protecting influence and brands (celebrity, political, commercial). The democracy of access became more of a threat to be managed than an asset to be cultivated, and the algorithms worked harder and harder to drive engagement and advertising revenue.

And so, we move on.

There isn’t a replacement for Twitter, nor I think is there ever likely to be. Instead, I am going to talk about the platform/network that I (and many other people who used to use Twitter a lot) have moved to over recent months and weeks. But before I do so, I think it’s really important for everyone to understand that it is not “the new Twitter”. If what you want is a new Twitter, you’re out of luck. We had that thing, and it’s done. Maybe something will appear in the future which will see politicians and celebrities put themselves out there in a pseudo-personal (and occasionally genuinely personal) way, but also maybe not. If you want to seek the attention of politicians or celebrities, then Twitter is still the place to do that. Twitter is likely also to remain the home of a lot of brands’ official social media, although perhaps following some of the recent (and very funny) exploitation of Twitter’s $8-blue-tick access, brands will instead seek to delegitimise what remains of Twitter for their own protection.

If you use an algorithmic feed for Twitter (either through its website or the official app), you may not have noticed, but a lot of your more personal contacts have just vanished: either they stopped tweeting (as I have), or they’ve completely deleted their accounts. Some are abandoning the whole idea of social microblogging, while some are moving to a new platform.

The new platform — which is an open-source application layer on top of an open-source protocol that has been in existence for many years now — is generally referred to by the name of the application layer: Mastodon. At a technical level, this uses a protocol called ActivityPub to link together a network of servers or instances, each of which has its own administrators and moderators, to create “an ensemble of federated servers”, dubbed the “fediverse”. So far, this all sounds like something Star Trek nerds would create, and I’m here to tell you that’s exactly what it is. Nevertheless, understanding how it works is very important to two things:

1. Understanding how you might use it, should you choose to join, and
2. Understanding that — once again — it isn’t Twitter.

Key to this understanding is recognising that there is no central control or management of Mastodon. While there are certain people who have been instrumental in developing the software and the protocols, anyone could take the existing source code, change it to do something different, and then set up their own instance and attempt to connect it to other instances. If successful in federating with other instances, then that new, different software would be part of the fediverse. On the other hand, if someone used the exisiting software or an existing instance to start hosting a whole bunch of problematic content, and didn’t adequately respond to concerns raised by administrators or moderators of other instances, they might find (and indeed have found) their instance defederated from the rest of the network.

This kind of structure is literally what the Internet is built on. The existence of megalithic centralised services that many people now consider to be “the Internet” is a weird inversion of the protocols and structures they use. For many people, if you were to block access only to the servers of Google, Facebook, Twitter, Instagram, TikTok and YouTube, they would believe that you had turned off the entire internet. And yet, that’s only 6 sites out of a global network of computers capable of connecting billions (or, with IPv6, trillions of trillions of trillions) of different servers and devices.

Given that centralised control is such anathema to the structure of the internet, it’s remarkable how dominant it is. Should you choose to venture into Mastodon, I warn you that you will need to adapt to the absence of such centralised control. It’s going to feel very, very strange at first. But I also commend to you that it is worth it.

To that end, I’m going to make the rest of this piece an introductory guide, so that people I know on Twitter but who haven’t yet explored Mastodon, can hopefully have an easier time of making connections in a new space.

Step 1: Find an instance

I found this to be the most unfamiliar and potentially confusing step. It’s important, but it’s also not the end of the world if you mess it up, so allow me to give you some guidance on what to look for.

Your instance is like your “home”, but unlike the real estate market, you’re not locked in with huge costs involved if you buy the wrong house — you can freely and easily move from one instance to another. You can also set up accounts on multiple instances, but I suggest this is not how most people will want to use it at least to begin with. Your initial identity on Mastodon will be a combination of your username and the name of your instance: for example, you can find me on Mastodon at @raaahbin@eigenmagic.net

It is good, if possible, to find an instance where you will have some natural connection or something in common with other people on that same instance. There are ultra-general instances (such as “mastodon.social”), which I would recommend against if you can possibly find something you identify with more closely. For example, “ausglam.space” is an instance dedicated to users who work or have interest in galleries, libraries and museums in the Australasian region. “prf.me” is an instance for people with interests in fragrance and perfumes. “meow.social” is intended for furries. You name it, there is either already an instance with relevance to a particular group of people, or there is likely to be in the very near future. The one I signed up to was simply recommended by a friend, and seemed like a good place to start. If in doubt, smaller (within reason) is generally better, because then the local timeline is more accessible and the admins and mods will likely be more responsive.

If you download the official app, or go to https://joinmastodon.org/servers, that will help you find servers that have given certain commitments regarding availability. There are also a number of directories of Mastodon instances — for example, https://instances.social/

Given the rate at which things are changing right now, those directories may be either overwhelmed or out of date or both. Don’t get discouraged if you find a cool-sounding server on a list, only to discover that it is not working or closed to registrations. There are plenty of other options.

And remember — if you change your mind later on, the ability to move and take all your follows with you is built in to the platform. So don’t stress too much.

Before signing up to an instance, one of the important things is to make sure you are happy with the rules of that particular instance. You will be expected to abide by the rules, and you can expect the admin or moderators of the instance to enforce them.

Rules of instances, and their enforcement, are crucial to the federated nature of the network. Each server admin may be called upon to determine whether they wish to remain federated with any other instance, and if an instance becomes a hotbed of problematic content, they probably won’t. There has been a bit of furore lately over the “journa.host” instance, where a whole bunch of journalists signed up and then simply behaved the way they always did on Twitter (such as posting a whole bunch of stuff with no content warnings and claiming they were upholding their journalistic integrity by behaving like inconsiderate dickheads)… and now that instance has largely been defederated from other instances on the network. So none of the rest of us have to put up with their bullshit on our timelines.

Step 2: Get an app

There are a number of really good apps for Mastodon already, and I expect there will yet be quite a few more. You can just use the website on your computer (many people do), but there are also good mobile apps in addition to the official app — Toot! and Metatext on iOS, Tusky on Android, as well as desktop apps (I use Toot! on mobile and Mastonaut on Mac).

Step 3: Get to know how Mastodon works

As I mentioned before, it’s not Twitter. It isn’t even really social media, in the generally accepted use of the phrase — rather, it is a social network. If you just try to use it like Twitter, you may be frustrated or disappointed or both. However, if you take a little while to learn to use it in its own way, you may find the experience far better than Twitter. I certainly have. I realised after about week of actively using Mastodon that I’d barely looked at Twitter (previously I would have looked at Twitter many times per day). I then spent five minutes looking at Twitter… and immediately hid it again. I haven’t yet deleted the apps or my account, but I’ve mostly abandoned them. I’ve already put the Toot! app in the spot on my phone screen where Tweetbot used to live.

It’s worth spending some time just browsing your timelines and getting to know people who are out there. The “follow” mechanism is similar to Twitter, in that it makes someone’s content appear on your Home timeline. If you see something good and wish to show the author that you like it, use the favourite function. If you see something good and wish to share it with your followers, use the boost function. But — here’s the important thing — no one else sees your favourites, and faves don’t do anything to the algorithm that drives content, because there is no algorithm that drives content. If you are looking at your Local timeline, that’s a straight time-based feed of content people are posting on your local instance. If you are looking at the Federated timeline, that’s a straight time-based feed of content people are posting on your instance and every instance it is federated with. If you are looking at your Home timeline, that’s a straight time-based feed of everyone you follow (including content they boost, when they boost it).

Those who are accustomed to Twitter will note the absence of the Quote Tweet function. This has been a conscious omission from the Mastodon platform, to avoid the toxicity associated with much of the use of that function on Twitter. However, there is an effective way to draw attention to someone else’s comment in the context of what you have to say about it: reply to someone’s “toot” (that’s the name given to a single post on Mastodon), then boost your own reply (that makes it visible in your follower’s timelines just like any other content you boost, but retains the context such that your followers can also go see the toot to which you were replying).

Step 4: get to know — and use — content warnings

One thing you will notice if you’re coming from Twitter, is that the “content warning” (or CW) feature of Mastodon is extremely widely used, and in many cases will be required by the rules of your instance when posting certain content. It is not my place to talk here about the enforcement of CW rules in cases of power imbalance (there have been cases where white folks have insisted that POC hide discussion of racism behind CWs because talking about racism makes white people uncomfortable… and honestly, that’s some bullshit). I’m just going to say that if you, like me, exist largely on the up-side of societal power imbalances, then you should take care to abide by your instance’s rules regarding CWs.

Mastodon has been built in large part by marginalised people — trans and gender diverse, LGBTQIA+, neurodiverse, etc — so it’s not surprising that structures to help protect marginalised people are baked in.

If you’re about to tell me how you aren’t a sensitive little snowflake and you don’t want every second post hidden behind a CW, then good for you — you can just set your own account to auto-expand CWs, and hey presto! CWs just disappeared for you, but not for people who might actually have valid reasons to want certain content not to display by default.

If you’re going to post sensitive or potentially triggering content, it takes all of two seconds to click “content warning” and put your toot behind it.

As an extension to the above, accessibility is also very important: when you post an image, take a few seconds to add an image description to it (again, this is built into the mechanism of composing a new toot).

Step 5: build your filters

While Mastodon may be inherently superior to Twitter at a structural level, that doesn’t mean it’s perfect — it’s still content generated and moderated by humans, and you’re not going to want to spend every day looking at every single thing that the people you follow might share into the timeline. If you go to your instance’s website, log in, and go to your account Preferences, there is an entire section called Filters, which can be used to filter out certain content. For example, if you want to avoid reading any mention of what some people call “The Bad Show”, you could add a filter to remove any toot mentioning #QandA. You can set filters to remove something altogether, or to hide it behind a content warning, or to hide it in certain contexts but not in others.

Step 6: have some patience

The down-side of this not being centralised is that individual servers may sometimes struggle, especially in times of major change (such as millions of new users coming into the network). While the protocol and overall structure are very resilient, if your instance goes down (or won’t immediately approve your sign-up request because they just got a deluge of applications)… then maybe just wait a while. And if the instance you chose either never lets you sign up or is always having server problems, then just sign up on a different instance. You could even spin up your own instance if you are so inclined, but for most users there isn’t likely much benefit in doing that, unless you have a specific ready-made group of people who will want to be on (and whom you want on) your own instance.

Also, when things like software updates happen, they might just need to take your instance offline while that happens. Don’t panic. Follow the account of your instance admin, so you know when such things are happening.

Hopefully that is enough to get you going on Mastodon. Much more has been written by people with way more expertise than I. If you want more info, a good place to start looking is the https://fedi.tips/ website.

If you do sign up, feel free to look me up (my account address is given up there in step 1). I’m cautiously hopeful that Mastodon may yet bring a lot of the good quality connections that Twitter facilitated, while remaining structurally resistant to the rot of centralised control and algorithmically driven content. After Facebook and Instagram and Twitter and the like, I think we’ve given the algorithm more than a few opportunities to get it right. Maybe this time, by putting implicit trust in the more-often-than-not good nature of people instead, we could actually end up making something better.