Maryam, the author of this article, is an undergraduate student pursuing a BS in Computational Physics. She has served as a Student Voice Leader at her college for two years and has been actively involved in academic research and computational physics projects. Her work includes research on the influence of aircraft geometry on Radar Cross Section (RCS), developing a 2D planetary simulation to track the speed and distance of planets, and currently working on a 3D projectile motion project. She can be reached at email.
We are more connected than ever. We can interact with people without physical proximity, cross borders without leaving our rooms, and communicate with people we would probably never meet otherwise.
There is also a certain freedom in being behind a screen.
You don’t have to deal with the immediate anxiety of pleasing someone face-to-face, and the distance can make it easier to express yourself as you actually are.
You would think that this unprecedented access to people from different backgrounds would make us more open, empathetic, and understanding.
Research gives us some reason to expect that. A 2021 study on political discussion found that talking to people from different groups can be associated with lower hostility toward those groups.
If contact across differences can reduce hostility, then a generation with more opportunities for that contact than any before it should, in theory, be becoming better at understanding one another.
Yet our digital world often seems to produce a very different kind of interaction.
We don’t simply encounter people who are different from us. We encounter communities built around those differences. We see arguments between political groups, religions, nationalities, genders, and ideologies.
We don’t just disagree with people anymore; increasingly, we categorize them.
Us. Them.
So, How Does That Happen?
The answer may be as simple as your feed.
The problem is not simply that an algorithm shows you something hateful. The more important question is: what happens when an algorithm learns what makes you feel that you belong?
The answer may lie in tribalism: our deeply rooted need to belong to a group.
Humans naturally categorize themselves and others into groups. We look for people who share our beliefs, experiences, and values. Belonging gives us identity. It tells us who we are.
Algorithms are exceptionally good at identifying those patterns.
You interact with certain content. You stop at certain videos. You agree with certain opinions. You react more strongly to certain arguments. The system learns from those behaviours and uses them to decide what to show you next.
It doesn’t have to tell you, “Join this group.”
It can simply keep showing you the things that make the group feel familiar. Then it shows you the people outside of it.
At first, the difference might be harmless:
“They think differently.”
Then it becomes:
“They don’t understand us.”
Then:
“They are against us.”
And eventually:
“They are the problem.”
When Differences Become Identity
The algorithm doesn’t necessarily create the original disagreement. It can take something that already exists—an insecurity, a political belief, a cultural difference, or a personal grievance—and repeatedly place it in front of you until that difference becomes central to how you understand the world.
You might think:
“I’m not that easy. You can never convince me to do something like that. I can think for myself.”
And that’s exactly why this is more complicated than simply asking whether an algorithm can convince someone to believe something.
You may not be able to convince someone directly. But if a person is repeatedly exposed to the same interpretation of events, they can eventually convince themselves.
The important part is justification.
You feel that you are protecting yourself. Protecting what’s dear to you. Protecting your people.
But what you’re protecting can become your very sense of belonging to that group. And eventually, it is no longer simply you defending your beliefs. It becomes almost everyone in your group.
Their beliefs become your beliefs. Their enemies become your enemies. Their suffering becomes connected to yours.
There are no longer simply you and them.
It becomes us versus them.
When Belonging Becomes Survival
And once belonging becomes connected to survival, the other group doesn’t have to be personally known to you.
They only have to represent them.
A fictional example of this can be found in Attack on Titan.
If you haven’t heard of it, the story is largely about a race of people being haunted for the crimes of their ancestors. The main character, being one of them, eventually decides that genocide is the solution. He forms his own faction and goes through with it.
Was genocide the answer?
No.
But would a 15-year-old who watched clips and edits of nationalist characters necessarily think the same?
Absolutely not.
You can find a large community of people who justify the genocide in the story—the deaths of around 80% of the world’s population in its fictional world. People who went against the genocide can be called traitors.
Characters such as Floch can be defended through arguments like:
“He’s understandable.”
“He was scared.”
“If you were called a monster all your life, you’d do the same.”
But fear isn’t an excuse to kill billions.
The people who are killed in a war as collateral damage have had zero control over the decisions that led to their misery or your misery.
Yet when someone becomes convinced that the other group is responsible simply because of who they are, individual responsibility disappears.
“They should suffer because of who they are, for what I did because of who I was.”
There is no longer a distinction between the individual and the group.
You become what you hate, and you feel justified about it.
So, you think you’re right.
You think you’re better.
And that’s a fictional example of it.
But What About Everyday Life?
People watched that story and decided that the genocide was justified. What happens when the same process isn’t concentrated in one fictional story, but happens gradually through the content someone consumes every day?
One post isn’t going to transform someone’s worldview.
Neither is one video.
But algorithms don’t need one piece of content to do that.
They can give you another video after the first one. Then another argument. Then another person saying the same thing. Then a community where everyone agrees with you. Then examples of the other side behaving exactly as you have already been taught to expect.
And because you keep interacting with that content, the system keeps learning.
The more you engage, the more precisely it can reproduce the environment around you.
Eventually, you aren’t simply being shown information.
You’re being shown a world.
A world where your group is understandable, threatened, and justified—and their group is increasingly difficult to understand.
The Power of “Us Versus Them”
That is the power of us versus them.
The algorithm doesn’t have to invent the division. It only has to learn how to keep you inside it.
And the more convincing that world becomes, the less someone needs to tell you what to believe.
You begin to convince yourself.
And if us versus them can become this convincing, if belonging can become so deeply tied to identity that people can justify almost anything to protect their group, then what about the people who refuse?
Who are the people who betray their own?
Who are the people who break free from the us and choose them?
Or sometimes, choose themselves?
