How to decentralize social media

Social media needs to become decentralized. If you’re reading this, you probably agree.

In my mind, the major threat to decentralized social media is simple; huge incumbent platforms have network effects that are almost impossible to overcome. These platforms are the opposite of everything we want; they are centralized, for-profit, opaque, and enshittified. We’re a million miles away from decentralized social media today, because these platforms are are huge and real:

For decentralized social media to matter (i.e. make a difference in the world), it needs to grow. I believe that growing Mastodon by attracting Twitter quitters is the best way to do that. In that essay, I describe growing Mastodon as a first step, and a way to generate enough scale in the fediverse to help it take off. A big obstacle to that was the challenge of getting started. So I helped create http://SpreadMastodon.org, a portal to join the fediverse that’s super easy to use.

Some people have expressed concern about a design choice we made, which was to direct people to mastodon.social and bypass the step of choosing a server. We did that for a few reasons:

1. Server choice is a major obstacle to growing the fediverse.

We are losing *so many* prospective users to server choice. I only have anecdata and wish I had more, but server choice is easily the most common complaint that I and many others have seen. Just recently, a smart tech professional emailed me his whole list of hobbies and interests, saying he was “at a loss” for how to join Mastodon and wanted my help picking a server.

2. Server choice is confusing, and we as a community haven’t fixed that.

Here’s my summary of the best guidance we give to people picking a server:

Pick a server based on your interests (or your geography (or your demographic)), and your choice doesn’t really matter except in these ways it kinda does, and you can totally switch later except for these things you can’t, and if you pick the wrong one the admin might disappear and you have to start over.

I’m caricaturing for humor, but only slightly. The point is that a better, cleaner, easier server selection flow has yet to be invented.

3. Server choice can have big downsides, and we as a community haven’t fixed that.

The approach of “just pick a server” disrespects the time and investment that people make in joining our community. If we care about users’ experience, server choice matters, and in ways not easily captured by a listing. Admins have sometimes lost patience and shut down (or threatened to shut down) with little notice, forcing users to move and laboriously back up or lose their history. Moderation and server-blocking can be over-used, such as in a recent discussion of blocking hachyderm because it disallows fundraising (edit: underscoring my point, that discussion appears to have been based on a misunderstanding). Plus there are not-yet-solved technical issues such as the challenges of migration and people on small servers not seeing all replies or post engagements.

4. The fediverse isn’t likely to become unduly centralized.

I 100% agree that keeping Mastodon and the fediverse decentralized is an important protection. If we reach a point where undue centralization became an issue, or seemed even foreseeable to become an issue, I’d focus more on it. But for now, the present reality of centralized social dwarfs that possible future risk.

Also, nothing is forever, including the design of current tools and apps. For example, SpreadMastodon may well evolve to a “round robin” that directs users to one of a small handful of servers selected for (among other factors) their stability and broad federation.

5. The real problem is already-existing centralized social.

To my mind, anyone who wants to decentralize social media has a clear shared problem to solve. That problem is the size of the left-hand bars, not the temporary make-up of those on the right:

To decentralize social media, my suggestion is that we work together with a primary goal of growing the right-hand bars. I‘m confident that the fediverse will remain more than sufficiently decentralized to solve the problems we want to solve. If at any point I’m wrong and fediverse concentration becomes an issue, I’ll be here to help address that.

How to further decentralize the fediverse

I’m incredibly grateful to all independent server admins; without them, there would be no fediverse worth building. So to some extent, anything that supports support server admins is helpful to fediverse decentralization. Beyond that broad concept, here are some specific projects that would support fediverse decentralization:

• Work on making server choice matter less, by addressing the issues in point #3 above such as:
  – Technical issues such as reply visibility and migration
  – Admins’ ability to make binding promises regarding the user experience they provide
• Build and test an improved server-choice flow until it has low friction, then share it for others to use
• Build and market a portal for new fediverse users, but direct people to other servers
• Build great ways for people to pick a server after they’ve had a rewarding experience on the fediverse and want to stay

I’m glad that we share the goal of decentralizing social media. We finally have an opportunity to achieve that, but I believe it won’t last for long. If we work together, I believe we’ll be able to seize the moment.