What would Twitter be if it adopted Wikipedia’s politics?
[First published on OpenDemocracy.net]
I’ve been mastoing. Actually tooting? In… Masterdonia? Ok, what I’ve really been doing is using mastodon.social (that’s the URL), a potential Twitter alternative – the community hasn’t quite got the lingo down yet though. It’s familiar, because it’s similar to Twitter, but at the same time, all the percolating aggression and designed-in conflict is absent. It’s youthful optimism to Twitter’s embittered decrepitude.
The Mastodon.social mascot: will you get to know this mammoth?Abuse of social media is a characteristic of the modern authoritarian – it gives the ability to speak directly to supporters without the context provided by a professional media. Meanwhile, Twitter is having plenty of corporate problems of its own.
Could now be the time for a cooperatively run, open source alternative to Twitter, perhaps run along the lines of Wikipedia? Owned by everyone, responding to community concerns: tools to defend against trolling, manipulation by Russia or to help stem the tide of fake news.
I’m going to suggest that the problem is not a technical one, but one of finding the right kinds of social institutions. But, before we get to that, you’ll have to come on a journey with me into the world of Mastodon, to venture into the (relatively nerdy) ‘fediverse’…
Federation: a technical approach for the free web
Mastodon.social is going to take a moment of explaining. Three paragraphs of explaining, as it turns out. The underlying concept is a federated alternative to Twitter, which means that there is no single authority that runs the system. Anyone with tech skills can set up a server, install the Mastodon software and have their own website (Mastodon is open source, meaning that you can copy and read the code for free). They can get their friends to sign up and exchange messages with them in much the same way as on Twitter. If they want to, they can connect their Mastodon website with someone else’s. This system means you can ‘follow’ and ‘@ message’ users on the other server.
So you end up with a constellation of privately owned servers. There is no central entity running things that can censor, or, for example, run a secret algorithm that might be disproportionately showing people ‘fake’ news stories.
The federated server model: Numerous, private servers, run by the community to their own tastes.
That ability for servers to communicate is crucial. Just like the email system is designed so you can send email from, say, Gmail to Hotmail, Mastodon users are able to communicate across systems owned and run by a diverse group of individuals or companies. The interconnected web of servers is sometimes called the fediverse or the ‘federated social web’ (see a complete list of sites linked to the system). This is the idea of federation. Mastodon.social is one website running the mastodon software, and it connects to lots of others.
Around November 1st 2016 I was happily tooting away on mastodon.social when things started to get exciting. More and more people started arriving. People with lots of followers started Tweeting about mastodon.social; lots of enthusiastic new mastoers tooting away. I’d been on Mastodon.social for all of the three days, so I took it on myself to show people the ropes, say hello and welcome them to the neighborhood.
Mastodon.social in action.
An Anonymous Twitter account with 10 million followers signed up, and digital privacy activist Aral Balkan began using the site, bringing the community around him to the site as well. The service began to falter because of all the traffic. When I got up the next day, Mastodon.social was full of Dutch people, Mastodon.social was trending on Dutch Twitter.
It felt a bit like the fabled ‘hockey stick moment’, when an app breaks through (so called because a graph of user numbers looks like a hockey stick – suddenly trending upwards).
What rules do we play by?
Then, a question came up: what are the terms of service? I thought that this was a perfect moment to discuss what Mastodon was about. Is it trying to be a Twitter-killer? Or is it actually not helpful to compare it with Twitter? How ought it to be funded, if at all? I have to report that most people were more interested in boxing off a few pragmatic concerns than embarking on a philosophical journey about what Mastodon could be.
And this is the moment where Mastodon began to exemplify a classic Open Source question: mainstream adoption. I was in a minority camp. As I’ve said, I would love an open alternative to Twitter. That means having something that a non-tech-person could understand. I believe you have to sand down some rough edges off the complex, federated model described above to make it usable to a general audience.
In the other camp, people wanted to keep strictly to the federated model. They either believed a federated model could attract mainstream adoption, or didn’t care about mainstream adoption.
What do I mean by sanding down the edges? To give one example, you cannot search across the ‘fediverse’. Even though you can @message anyone, or follow anyone, if you search for #presidentialdebate, you won’t see most of the messages that are stored on other servers. To me, that is a very important feature, accessing a global debate is part of the appeal of Twitter.
That’s an example, but it actually doesn’t get to the nub of the issue. The issue is: you need to understand federation before you can understand what Mastodon offers you. It’s not a nerdy detail, it’s something you have to understand before you are going to sign up. Where do you even sign up? The aspiration is that there will be multiple servers with different URLs, each of which will give you a slightly different window on the Mastodon world.
It took me three paragraphs to give a picture earlier, and that skated over quite a few technicalities. Three paragraph explainers are not a good sign that an idea is accessible.
As mentioned, technically, email is federated, so everyone has been using federated services for years. But email is a) super weird in many ways and no guide in understanding user behaviour, and b) private, so many of the most confusing issues don’t come up.
I understand what the other camp was saying though. The argument is that if you remove federation what you have is Twitter, which already exists.
Open source code, good visual design and…good institutions
That’s not quite true though. Let’s look at a deeper level. The problem is not federation versus centralisation, but a question of the kind of offer you can make to users. Twitter, it seems to me, ought to run on a Wikipedia style model because I want a system that cares intrinsically about being fair and legitimate.
Some key values here are that Wikipedia:
- Runs for the benefit everyone, without advertising
- Is transparent about how it works, including open sourcing code so anyone can see it
- Provides easy access to content that its users created (including bulk access)
I don’t mean a Twitter alternative has to work exactly like Wikipedia, but it’s a good steer. I sense that a lot of people feel the same, and would greatly value a service that offered those things. The important thing here is that the ‘institution’ supporting a website (be that a company, a foundation, a community of coders, etc) is part of the offer you make to users, not some ignorable bolt on – convincing people of that is the only way to an ethical web.
There are three parts to delivering such a thing, one is clearly the tech part, and that’s being delivered extremely well in many open source projects. Another is the user interface – open source is not so hot on this one (it’s often ugly and difficult), though Mastodon has made user interface a key part of its offer, and is succeeding because of it.
The third part is is that crucial institutional bit. Mastodon is doing a really good job of attracting donations – it has over 100 people donating more than $500 a month and it’s going up all the time. But my experience asking questions about Terms and Conditions suggests that this part of the project could be running better – that the institutional design isn’t quite there yet.
The institutional part does some of the heavy lifting for the tech part. Instead of trying to find software that is federated and usable by normal people (probably impossible), you ought to be able to come up with a version whose legitimacy stems from the organisation that runs it even if it isn’t completely federated. Tempting as it is for developers to see ‘ethical Twitter’ as purely a technical problem, that’s never going to work.
The next step toward a community owned Twitter alternative
What’s needed is to reach out beyond the nerdy fediverse and involve people with institutional expertise: lawyers, activists and volunteers who can help build that institutional foundation which can take technical capabilities and transform them into functioning social entities. If you are one of those people, perhaps you should get involved at mastodon.social.
Such systems will not automatically prevent abuse by authoritarian politicians, nor can they immediately distinguish fake news from real. What they can do is provide transparent systems where all the participants are aiming for fairness not profits.
If we are very lucky, when politicians see what open source and cooperative institutions can do they might stop being so hypnotised by the market, and realise the plurality of approaches that are appropriate in the digital era. Instead of a handful of corporate giants showing you endless spammy adverts, the web might get a chance to be what it always wanted to be: a platform for democracy, personal agency and empowerment.