You Don't Love Them — Your Algorithm Trained You To
You've never spoken to them.
But you know the way they hold their phone when they're filming. You know which filter they use. You've noticed that the timestamps on their posts are usually late — past midnight. You didn't go looking for any of that. You didn't follow them first. The feed kept returning them to you, one surface at a time, until the details started accumulating. A comment on someone else's post. A video you didn't click but watched to the end. A face in the corner of a reel, then again at the top of your scroll, then again in a thread you weren't even in.
You started thinking about them between sessions. Started wondering if this is what it feels like when something is meant to be.
It isn't. What you're feeling is real. But you didn't generate it. Your feed did.
Zajonc, 1968 — The Finding That Matters Here
Robert Zajonc published one of the most replicated findings in social psychology in 1968. The study was simple. Subjects were shown images of unfamiliar faces at varying frequencies. Some faces appeared once. Others appeared dozens of times. Subjects were then asked to rate how much they liked each face.
The result was clean and uncomfortable: the more often a face had been shown, the more subjects reported liking it. Not recognizing it. Liking it.
Zajonc named the phenomenon the mere exposure effect. Repeated exposure to a stimulus — any stimulus — produces genuine feelings of familiarity and warmth. The nervous system runs preference updates below conscious awareness. Familiar becomes good. Good becomes safe. Safe gets coded as something closer to attachment than it has any right to be.
The finding has held across five decades of replication. It works with faces, words, shapes, sounds. It works when subjects don't consciously remember being exposed. It works when the exposure is subliminal — too fast for the conscious mind to register, not too fast for the nervous system to log. You don't have to notice something for it to move you toward it.
A 2026 study on algorithm-mediated social behavior confirmed what Zajonc's original work had only implied in analog form. Recommendation systems weren't just showing you content you might like. They were producing feelings of connection through controlled, repeated exposure — and the users experiencing those feelings described them as destiny.
How the Algorithm Manufactures the Feeling
Recommendation systems don't understand psychology. They understand behavioral signals. They track what produces return visits, what generates rewatches, what keeps a user in the app for the next minute and the minute after that.
What they discovered — through optimization, not theory — is that certain faces keep users in the app. Not because those faces are objectively more attractive. Not because those users post better content. Because the algorithm had already shown you that face three times this week. Familiarity is doing the work. The system found the mere exposure effect before researchers thought to study it in this context.
The mechanism is straightforward. You open the app. The recommendation engine surfaces a profile — someone you don't follow, someone with no obvious connection to you. Maybe you watched a video for eight seconds before scrolling. The system logs that. Next session, the face comes back. You scroll past it. The system logs the pause. It comes back again. You click. You spend forty seconds. The system has now confirmed this face is performing above baseline for your engagement profile.
It will keep showing you that face.
Each appearance is another exposure. Each exposure produces familiarity. Familiarity produces warmth. Warmth, when it builds without a clear source, gets interpreted as something the mind has no better word for than connection. You feel like you know them. You feel like there's something there. The feeling is accurate. The system made sure of it.
The product layer is invisible. You experienced the outcome. The mechanism stayed hidden inside the recommendation engine.
The "Sign" the Algorithm Built
The 2026 study found something specific in the qualitative data. People who experienced algorithm-mediated attraction didn't describe it as "I've seen this person a lot." They described it as fate. As cosmic alignment. As the sense that the universe was surfacing this person for a reason.
That framing isn't delusion. It's the mind doing exactly what it's supposed to do: generate an explanation for an emotion. You feel warmth toward this person. You feel familiarity. You feel a pull that doesn't have an obvious cause. Your cognitive system scans for an explanation and finds none that maps to ordinary social experience — you haven't met, haven't talked, have no mutual history. When the rational explanation isn't available, the mind reaches for a larger framework. Significance. Fate. A sign.
The emotion is real. The explanation is wrong. The actual cause was sitting in the product layer the whole time — systematic feed optimization that had been running on your behavior for weeks before you consciously noticed the face at all.
Repetition created familiarity. Familiarity created warmth. Warmth without a visible cause became destiny. The feeling of meant-to-be was manufactured before you opened the app that morning.
What You Can Actually Do With This
Mute or block their content for 14 days.
The mere exposure effect runs in reverse. What you stop seeing, you stop craving. The warmth response that exposure built, non-exposure will dismantle. This isn't suppression. It's removing the input signal and watching what happens to the feeling without it.
If the feeling dissolves in two weeks, you have your answer. The algorithm built it. It can't survive without the feed reinforcing it. What you were experiencing wasn't connection — it was the output of an optimization system that found your attention and kept pulling it back.
If the feeling doesn't dissolve — if it persists outside of feed reinforcement, if you find yourself thinking about something specific and particular about this person that exists independent of their content — then you have cleaner information. You want something that survives the absence of the mechanism. That's worth knowing. It means something different.
Either outcome gives you more to work with than the feed ever will. The algorithm has no interest in helping you distinguish between manufactured familiarity and something real. It optimizes for time in the app. What you do with the feelings it generates is not its problem.
Your nervous system wasn't broken. It was working exactly as designed. Someone else just got access to the controls.
The feeling was real. The feed made sure of that. Real is not the same as true.
Follow The Shadow Self Media on TikTok · Instagram · Facebook