Social media: designed to be bad for you & for society

After writing this, I wasn’t happy enough with the result to post it. However, reading spkoc’s comment on a recent COVID update made me think that I should share it because it seems as relevant as it did when I typed it out.

This post is my second attempt to explain and check the worry I feel about social media. The feedback from the first one told me that instead of clarifying my thoughts, I only managed to convey a vague sense of unease. Let’s give this one more shot then.

Mainstream opinion of services like Facebook, Twitter, etc. is trending negative. It is said that, on a personal level, they’re addictive, wasting people’s time and harming their self-esteem. At the group level, they create echo chambers, fuel conspiracy theories, and even enable ethnic cleansing. But I think all of this misses a crucial point–that all these things are connected, parts of a single system, operating within normal parameters. And the negative effects I described? Merely expected by-products, like car exhaust or traffic jams.

This error leads us to underestimate just how much these platforms degrade global epistemic conditions. To illustrate my point, let me describe three mechanisms that I consider core to social media.

As Digital Nicotine

I used to smoke back in high school. It made me feel good–calm but awake–especially between classes when I worried about my grades and my future. Cigarettes also helped connect with other people. Smoking signaled that I was alright, that I wasn’t taking life too seriously and that I could be trusted.

Social media feels similar. Seeing photos of friends partying or people liking your post activates some ancient circuits in the brain. The ones that help us work together, build relationships with others, and track our position in hierarchies. The same ones which allowed our ancestors to form stable hunter-gatherer bands, which turned out to be such a good survival strategy that we’ve taken over the world. That’s why it feels so good to to consume all that the social web has to offer: cute babies, pretty bodies, and outrageous news.

It would never have worked out without smartphones. They freed users from stationary, beige boxes that had to be shared with parents or siblings. Suddenly, everyone had a private device, a little portal into the social universe, so they could check in whenever and wherever they were. It was handy, instantaneous relief from boredom, kind of like having a quick smoke.

As Video Game

In games, players often create avatars, join groups, fight monsters, and collect experience points. On social media, they create profiles, join echo chambers, fight people from the outgroup, and collect followers. Both are fun and engaging and difficult to put down.

There’s another genre of games that are hard to quit: slot machines. Casinos have greatly advanced the art of making them more addicting. For example, they play faster and with shorter breaks, increasing the number of games played per unit of time. That’s clever, but still within the realm of acceptable practice. But how about this piece of psychological black magic: slot machines simulate the player almost winning, evoking the emotion of “almost getting it”, tricking them to play another round. The only purpose of these techniques is to keep people playing, even if there’s no end-game.

In a similar fashion, the most important metric for a social media platform is engagement–how much time users spend actively using their site or app. To increase that metric, platforms often introduce features like videos or payments. But sometimes, they go for something a little more clever, like algorithmic feeds. These track a user’s interactions and use that to display content that’s likely to generate more interactions, no matter if it makes users feel good or bad. What matters is if the user will come back and play some more. Luckily for the platforms, there is no end-game.

As Attention Harvester

In “The Attention Merchents”, Tim Wu describes the story of modern advertising. It begins when one man, a printer by trade, noticed an opportunity in the daily paper market–if he could get people to pay him to print stories about their products, he could drop the price from 5-6 cents to just 1 cent. But to do that, he would have to ensure people were reading those stories. And he had just the idea to make that happen.

His paper printed scandals, crime reports, and sometimes complete fabrications. It was the first paper to hire a reporter dedicated solely to hanging around the courthouse and producing stories about all the wild and ugly cases that flowed through those walls. It worked so well, in fact, that the paper’s circulation exceeded that of any other local competitor. As others copied his model and achieved success, the advertising industry was born.

Over time, the model was applied to other media like radio, TV, and finally the Web. And while it was refined and expanded every step of the way, the core remained the same: capture the attention of an audience and sell it. There are three parties to this transaction: someone who wants to sell a product or service; an audience seeking entertainment; and the middleman who captures the attention of the second group and sells it to the first.

(This is where the phrase “If you are not paying for it, you’re not the customer; you’re the product being sold” originates from.)

Social media companies are the attention merchants. They’ve refined the harvesting process by collecting data about their users and using it to display ads that are most likely to engage them. When one describes this in a positive light, they point out that these are ads about things people really care about. But attention merchants are locked into competition over a finite resource. So there’s no natural limit to how far they’re willing to “optimize” their harvesting. That is, at least in Wu’s take, until the “product” revolts, like it has a few times before.

The Crux

This then is the crux of my worry: social media is an inherently exploitative medium. There’s a lot of systematic effort–researcher/​engineer hours–that goes into making it grab as much of people’s attention as possible at any cost. Because we ignore this point, we look at social media’s side-effects as separate, contained externalities, which in turn suppresses any society-level response to this problem. Until we find an alternative, basically a new set of schelling points for online presence and advertising, I expect these problems to continue getting worse.