The challenge with running a great internet platform is that there are no guarantees that you’re going to be able to keep it around just because it’s great.

In a world where venture capital often shapes exactly what the internet will end up looking like, folks who want to work outside that system are taking some giant risks.

All of this is to say that the best parts of internet culture are typically noble failures. And, looking in aggregate, can be hard to see the nobility in the failure.

I bring this up because I’m sad about a social network that just announced it was shutting down at the end of the year. That social network, cohost, faced long odds, because it worked outside the obvious systems. It wasn’t connected to the fediverse, or funded by a billionaire benefactor. It was a site where creative people could hang out, with no algorithm in sight. Worst of all, you had to actually work to discover people. Or maybe that was the entire point.

Cohost was built by something called the “anti software software club,” which calls itself “a small team of developers and designers who like sharing things on the internet.” In other words, it was a play for internet culture in its purest form, along the lines of the legendary MetaFilter, and shared DNA with LiveJournal, DeviantArt, Tumblr, and even MySpace. Plus it was hacker-friendly: You could modify CSS on a per-post level, which begat the trend of “CSS crimes.”

I couldn’t find a home for it in my digital life, but it’s not because I didn’t admire what it represented. We are creatures of habit, and if you can’t find the habit that makes sense for you, it makes it hard to keep with it. (I won’t lie that I had considered reviving ShortFormBlog on the network with the benefit of that custom-CSS feature.)

Cohost was clearly built by artists for artists, and auteurs don’t always do things because they make the most financial sense. It feels like a platform built because the idea of the platform was cool, rather than any bigger long-term goal.

But eventually, even with the cool, interesting platform, things simply may not add up. The most recent financial update from cohost had the toxic combination of increasing expenses and shrinking revenue. Their losses doubled in a single month, while their income fell by nearly half.

At first, cohost had the help of a benefactor giving the network runway to reach stability, but the limits of that arrangement started to show its limits earlier this year. And despite the network’s honesty and transparency with its audience, it just did not work out in the end. Some have already written useful critiques of what happened with cohost; Aftermath’s Gita Jackson, who had critiques of cohost’s business model but was nonetheless understanding of the gap its loss creates, had some particularly well-considered thoughts on the matter.

Cohost, despite its small size, had developed a rich culture of shitposting and “CSS crimes.”

“I hope people understand that there’s a lot to be learned from the failures of cohost if you strip away the fandom aspect,” she wrote on Bluesky.

One of the lessons, to be clear, is not that independent social networks don’t have a place on the modern internet. I think a look at the success that the fediverse and Bluesky are seeing points at the potential for social networks to not necessarily be tethered to corporate ownership or venture capital investments. I think cohost’s big challenge is that it was an island unto itself at a time when we expected our social networks to live bigger lives.

Way back in 2022, when cohost launched, Mastodon creator Eugen Rochko expressed disappointment that the fediverse’s decentralized DNA did not inform what cohost made:

Decentralization isn't everything. It's not the goal in itself. There is a long tail of product design decisions that make or break a platform (which is why I find it kind of silly that Bluesky spent 2 years arguing over protocol design, though that's a different story). But that point is that decentralization is the foundation for a viable platform. Without it you're doomed to repeat the path of hundreds of social media apps of the past … Just a different paint on the same exact structure.

In one sense, it’s hard to disagree with this stance seeing where things landed. (This was posted well before Mastodon and Bluesky received their post-Twitter glow-ups, by the way.) But I don’t think they were ultimately building a social network. I think it was an art project to see what they could get away with in the social media artform. That’s not a criticism.

I see a lot of my own work in what they tried to create. The willingness to not go for the massive payday in favor of something designed to be user-friendly? I get that. The decision to underplay the discoverability element that leads people to follow the same groups of people on every platform? Ditto. The embrace of artifice over convention? That’s something to respect. And, for a social network, the design of the site was fucking weird! That is something we don’t see enough of nowadays. Convention too easily wins.

It’s sad that cohost is leaving, but it’s because I see a weird creature of the internet that was unfortunately malnourished, never receiving the level of support it deserved. It takes one to know one.