It's Time to Ban Algorithmic Social Media

Okay, so another clickbait-y headline, but, again, another headline that I mean in all seriousness. It is a sad commentary on the world that the most commonsense items come across as the most radical. Easier to imagine the end of the world, apparently, than the end of Silicon Valley fscking people over.

I am not, by the way, advocating for the end of social media. Social media is just a product, like any other product. A fire extinguisher that puts out fires is a great little tool. A fire extinguisher that sprays gasoline on fires is less so. Social media that allows you to choose who to follow, shows you everything the people who you follow posts as they post it, and doesn’t push ads or material you have not explicitly signed up for is fine. Social media that pushes material into your feed that you haven’t asked for, and does so in ways that radicalizes people and encourages them to harm themselves or get ripped off is not so good. The latter is what we have, and it is time to do something about this.

The Guardian has an excellent piece on how Twitter (not X. If he can deadname his daughter, I can deadname his website) pushes far-right content on people. It is an incredible story, extremely well written and well researched story, so I encourage all of you to read it. The take away is that Twitter’s algorithm pushes far-right content at everyone. Even if you identify as a leftist, you still get mostly right wing, often very extreme nazi or nazi-adjacent content. If you are right wing or neutral, you get almost no left wing content and an overwhelming amount of far-right wing content. More than half of that content, for example, came from accounts that used hateful or otherwise extreme language. But its not just forcing Nazi-like content down the throat of users. Social media is designed to cause harm in other ways.

About ten percent of Meta’s ad revenue comes from fraud — and it is aware of this. Worse, if you click on a fraudulent ad, you are more likely to be served other fraudulent ads. The Meta algorithm pushes ads similar to the ones you have shown interest in and thus people who are victims of these ads are the ones who are shown more ads that want to hurt them. It is not just money, however. Meta knows that it shows content about eating disorders — to be clear, not content that talks about the damage of eating disorders, but content that could encourage eating disorders — to teens after they express dissatisfaction with their bodies.

The common denominator in all of these situations is the algorithm. The algorithm controls what you see on social media, and it is more and more attempting to turn you into a Nazi with eating disorders who is constantly being ripped off by scammers. These are broken products apparently designed to harm you. Since the harm comes from the algorithm, the solution is to ban the algorithm. Social media sites should not be allowed to push content into your feed. All content from everyone you follow should be presented to you. No ads into your feeds — ads should be in a separate section. Recommendations should be only about who to follow and not specific posts and, again, should be siloed away from the main feed. None of this is perfect, of course, but we need to stop allowing broken products to remain broken.

Social media algorithms are harmful products. They are designed to hurt people and society in order to put money into the pockets of their owners. Any functioning society would put an end to these harms, and reigning in the algorithms is the surest path to that outcome available. It only seems radical because we have spent the last twenty-five years pretending the Internet is a magical land. it is not, and the firms running it are just firms. They should be treated as such, not as semi-mystical oracles of the future.