The First Million

"As of the last hourly update, the greater fediverse has passed 1,000,000 accounts for the first time, with 1,015 active instances"

Of course numbers of accounts is not all that interesting, and following the usual 80/20 rule probably only 200,000 have any non-trivial level of activity. Some percentage of that will also be bots, but even if it's half then that's still massive compared to 2016 or any time previously.

In terms of user adoption 2017 has been the most successful year for the fediverse that I've seen. I've been running federated social network systems since 2010, beginning with Friendica (which was originally spelt with a 'k' rather than 'c'). For nearly all the time span between then and now the world of independent social networks was a small club of familiar faces. To be in that club you needed to be serious about remaining independent from the big commercial systems. Probably the number of active users before 2016 was in the tens rather than hundreds or thousands.

The growth of the fediverse didn't happen all at once. It came in waves. The biggest ones were in February 2016 and then again in April and October 2017. Considering that the previous world had been a twee one mostly comprising of hardcore Free Software supporters and veteran anarchists/punks it was a bit like having your server hit by a tsunami of recently ex-Twitter users. GNU Social, particularly with the Qvitter user interface, looked very similar to Twitter and so at the beginning of 2016 if you did a web search for something like "open source alternative to Twitter" then you'd probably find it in the top results.

Each wave has brought with it a different kind of culture, and the fediverse is now a much more diverse space than it was before 2016, which also makes it an exciting place to be. Roughly categorized, in 2016 the influx was mostly gamers (who somewhat clashed with the old-timer anarchists). In early 2017 it was mostly the LGBTQ community and in late 2017 it seemed to be mostly cartoonists and other artists.

Politics meets Technology

The reason for the renewed popularity of a system which isn't new and dates all the way back to the beginning of StatusNet in 2008 is not just a technocratic story of gradual improvements in software or the addition of new features, although some of that has of course occurred. It's at least as much about the social problems which Facebook and Twitter have encountered, the overall background of the rise of fascist politics and the reaction (or inaction) of sites like Twitter to that.

For Facebook and especially Twitter the introduction of algorithmic timelines helped to further exaggerate the bubble effect in which users only encounter people or opinions similar to those which they have encountered before. This also created further political polarization and fighting between rivaling ideologies, and by 2017 the level of toxicity in these big centralized systems was getting to an extent that many people were finding uncomfortable. Especially if you held minority views you could soon find yourself on the sharp end of harassment or dogpiling. The fediverse was a way of escaping from that, because it's structure is different from the centralized giants. It's not that everyone in the fediverse behaves well towards each other but that it's easier to avoid clashes and harder for bad actors to search for users to victimize.

So far the fediverse also has very little in terms of advertising or commercial activity. This means that the kinds of pressures which apply to Facebook and Twitter are irrelevant in the fediverse and the user experience tends to be freer of unwanted noise.


2018 and beyond

Will the fediverse continue to maintain its current level of popularity? Technology moves quickly and so there are no guarantees of anything, but I think the kinds of problems which Twitter is encountering are not the sorts of things which are going to go away easily and might ultimately be their downfall. I think they're inherent design faults for a socio-technical system which they can't do much about without breaking the business model. Even if they develop sophisticated AI to try to remove the bad behavior and "fake news" I don't think that will be very successful and I expect it will achieve the opposite of the intended result.