The Algorithm
Aside from the billionaires behind the oligarchy, we are constantly pushed, pulled and influenced by algorithms. It's a constant battle, and emotionally draining. There are, however, things we can do.
Algorithms on social media sites exist because of people. If it wasn’t for people providing data, algorithms wouldn’t exist. So we, you and I, are each part of various and sundry algorithms. When ideas with which we agree go viral, we sing the virtues of algorithms and when we disagree, we curse them.
A couple of weeks ago I watched a disturbing example of the underbelly of the algorithmic universe in which we live. A social media presence that I respect tried to bridge the divide between left and right. They made a mistake, in that someone who has just begun awakening from the MAGA disinformation cyberscape has to begin with baby steps. If such a journey is mined by FAFO, ‘I told you so’ barbs and spears being figuratively hurled by a roaring mob, said journey ends before it has even begun.
And that’s what happened. A few sharp tongued naysayers gave their “expert” opinion, the algorithm did its thing, and a stampeding herd of angry runaways followed.
Each social media site has a red silo and a blue silo. That’s how they drive engagement. That’s how they make money. Most of us know how it works, at least in theory. You tepidly support a viewpoint. Perhaps you even repost it, but you have a doubt or two. Then you see another post that’s firmly in opposition and your doubts increase. A couple of days of influencers vocally arguing against it and you’ve changed your mind.
It doesn’t happen if you’re ten toes down on something, only if there’s a doubt from the beginning.
If a post has strongly emotional content, it stands a good chance of going viral. That’s the problem. Whether an idea has merit isn’t important. Eliciting a strong emotional response is what makes the money machine work.
An article in Social Media Today had this to say: “...maybe there are benefits to implementing more stringent controls over social media algorithms, and limitations on what can and can’t be boosted by them, in order to address the constant amplification of rage-baiting, which is clearly causing massive divides in Western society. The U.S., of course, is the prime example of this, with extremist social media personalities now driving massive divides in society. Such commentators are effectively incentivized by algorithmic distribution; Social media algorithms aim to drive more engagement, in order to keep people using their respective apps more often, and the biggest drivers of engagement are posts that spark strong emotional response.”
That, however, is waiting for someone else to do something about it, waiting for some politician to pass a law or some bureaucracy to install a regulation. While we’re waiting for a hypothetical good guy in office to do something, there are small steps that we can take, that when aggregated, can make a difference.
Not everyone needs several hundred thousand followers on social media. If you’ve got that many, more power to you. If you don’t, then while you’re trying to get there, consider taking action in small groups to multiply your efforts.
For several years I have been a member of a group messaging room on Twitter. (I know, I’m supposed to call it X.) It’s set up so people can message posts they like to the group. Then the members can repost the ones that catch their eye. Reposting everything isn’t required, just pick and choose. (Unfortunately there were some who never reposted, some of whom were politicians added in 2023-2024. That doesn’t work very well.) This did help move the message, up until after the 2024 election, when a lot of people left the site. Unfortunately, right now it’s a ghost town. To be functioning at an optimum level, there should be from 10 to 20 active participants. There were several of these “rooms.” There need to be hundreds.
A couple of weeks ago I asked ChatGPT how to counter algorithms on social media, and surprisingly, what I’ve just described was one of the recommendations, with the added advice to scale it up. Now I’m old school. I’m rightfully cautious about depending on machine learning for anything of importance. At least at this current stage of AI infancy. I’ve seen this concept work, though. Leading up to the 2024 election, most of the handful of reposts I was getting on Twitter were from that group messaging room. Most my followers were only interested in getting more followers. That was my fault, as I had taken someone’s advice when I started and pursued the rapid goal of several thousand followers. But that's a different story.
From experience, using a group DM to suggest important posts to other members works. I can see where it would be useful in countering the negative effects of algorithms.
Another objective, particularly with something that causes you to be doubtful, is to slow the algorithm down. An emotion grabbing post is designed to trigger an instant response. That’s when it’s best to pause for a minute, fact check and verify, then, if reposting it is of social benefit, go ahead and share it.
Martial arts training builds a calm center and teaches grounding each action in that calmness. Allowing panic or rage to breach that calm can lead to defeat. Our battle is psychological, not physical. Finding the quiet place in the center of the storm is even more important. Our adversaries are not each other. This is not a left versus right issue. This is all of us against a divisive algorithm. This is, and always has been, working folks against those who would be a modern aristocracy, those whose goal is control.
All social media sites use some form of algorithm. Substack and Bluesky both have coded a much less invasive version. They are quiet ports in the storm.
For most of the other major sites, a battle ensues from the moment you first create a post. If you haven’t been aware that it existed, you now are. Being strategic will help us win.
Notes:
This piece really made me think about the fascinating but also kinda scary complexity of how these sistems operate; you’ve articulated the user-algorithm feedback loop so perfectly. It’s like when I’m trying to find new books online and suddenly my feed is only recommending similar geners, making me realize how easy it is to get stuck in one type of 'silo' even with something as simple as reading habits.