The Sameness of Social Media

Wednesday February 10, 2021 — Taipei, Taiwan

Every big social media company wants to be the same thing. The exact shape changes as time drags on — the past decade has been “stories,” ever since Snapchat, and it looks like the next decade will probably be whatever the generic term for what TikTok is doing becomes, but whatever the hot thing is, every company wants to do it. I think the pipeline here goes like this:

  1. Hiring engineers is expensive
  2. So you need to make a lot of money if you want to survive, in order to be able to pay salaries
  3. So you need a lot of users, in order show them a lot of advertisements
  4. So you need to copy whatever is popular, since that’s (definitionally) proven to be the thing that attracts the most users

The thing is, this sucks for people who actually use social media platforms! If you don’t like stories, tough shit, that’s what’s in right now. Facebook can decide that video is the new thing, and if you liked text, sucks for you, everything is video now.

I love small, idiosyncratic social media platforms like Special Fish. There are an infinite number of strange and unknown social mediums waiting to be explored. Here are a handful of points in this space:

I don’t think any of these are the “next big thing” that’s going to make a billion dollar company. If I was a VC, I probably wouldn’t bet on any of these ideas. But I think that all of these are projects that could easily make a wonderful social space for dozens or hundreds of people. Why isn’t that enough?

It seems to me that it’s because of the first point above — hiring engineers is expensive, and you need engineers (or, at least one engineer) to make a social media platform.This is how special.fish can exist — the person who made it was able to make it by themselves, so they don’t need to pay someone $150k/year or whatever to build and run it. But why? I don’t think it’s fundamental that putting together a social media platform for a few hundred people requires skills we think of as “engineering”.

Right now if you want to build a social space on the internet, you basically are limited to reddit/discord/slack/discourse/google groups/etc. The only real axes on which you have choices are:

I want to see a toolkit that lets anyone build a custom social space, with a much larger set of primitives to work with. I don’t think this is impossible, I think it just requires wanting it, and not shying away from giving users power to configure the experience that they have.