Why Do We Like Something So Much in the Digital Environment? Recommendation Algorithms, Dopamine, and Attention
Recommendation Algorithms and the Fragmentation of the Digital Environment
For a long time, we took for granted that, in the digital environment, recommendation algorithms did not decisively alter what we saw, and that when we talked about this environment, we were more or less talking about the same thing. That we shared a common space, even if we disagreed. That we saw the same things, even if we interpreted them differently. However, the contemporary digital environment is no longer organized as a shared space, but as a network of recommendation systems that filter what appears before each individual.
This transformation is not superficial. Much of what we see today is mediated by recommendation algorithms that select content based on our minimal interactions: how long we look at something, what we repeat, what we ignore. Algorithmic personalization not only organizes information; it organizes the experience. What appears before us is no longer simply what circulates, but what a system has calculated will best retain our attention.
For a while, the idea of a shared debate worked. Today, that intuition begins to fail. Not because words have disappeared—public opinion, discussion, exchange—but because the digital architecture that sustains our attention no longer distributes a shared experience, but rather multiple personalized journeys that barely touch.
Two people can live in the same city, go through the same events, and share similar cultural references, yet still encounter increasingly different content. Not because the common world has disappeared, but because daily experience is unknowingly organized around recommendations that reinforce what each person already watches, follows, or consumes.
Previously, even under rigid frameworks—religions, nations, parties, television—there was a shared cage. It limited and excluded, but in a collective and visible way. There was friction between stories, clashes between visions. The environment could be narrow, but it was common: what was shared did not guarantee agreement, but it imposed a foundation.
Today, the cage is no longer social in that sense. It is personal. It does not appear as a single framework within which we all move, but as an environment that silently adjusts to each person. From there, the world that appears begins to differ from the outset. We do not see the same things and then disagree about their meaning; we see different things from the very beginning.
That is why disagreement changes form. It no longer arises mainly from opposing positions regarding a shared reality, but from the failed intersection between experiences that barely coincide. We do not argue about the same thing from different viewpoints; we speak from environments that hardly overlap. And when that happens, conflict stops feeling like a difference that can be worked through and starts to feel like interference: something intruding where it shouldn’t and forcing a stop when what the environment has accustomed us to do is to keep going.
Closure of Meaning and Digital Attention
Something similar happens with meaning. Not as a profound truth, but as something simpler: the sensation that what appears fits us. That what we see confirms who we are and how we believe the world works.
Recommendation systems are designed to hold our attention. To do so, they show content similar to that with which we already interact. Over time, what interests us appears more; the rest, less. Our daily environment fills with what we already recognize, and the world we see starts to look coherent, aligned with our preferences.
And here is the decisive point: it is not that we come to know ourselves better. It is that the environment continuously returns the same version of ourselves. Repetition creates familiarity, and familiarity generates security. Thus, what was a momentary reaction ends up becoming the principle that organizes what we see.
In this way, the world begins to close without the need for prohibitions. Not because there is a lack of information, but because recommendations prioritize what already held our attention. Not because there are no options, but because the same ones appear again and again, reinforcing the impression that the world is exactly as we see it. Everything fits. Even what is missing.
The Matrix and the Myth of the Shared Illusion
To face that feeling, we typically resort to a familiar image. We say that we “live in The Matrix”, as if this coherence were a great collective illusion: a lie imposed from outside, a false world that hides the truth. The comparison is reassuring because it places the problem elsewhere.
But that’s where it fails. In Matrix there is still a common world, even if it is artificial. Everyone sees the same thing and everyone could wake up at the same time. The closure of meaning that we experience today works differently: it does not appear as a single lie, but as a different coherence for each person. It does not produce ignorance but confirmation. And there is no need to hide an exterior: it is enough that, little by little, leaving loses its meaning. Not because that world gives us everything we want, but because it explains everything.
The Truman World and the Personalized Experience
More than The Matrix, the contemporary problem is much more like The Truman Show (Peter Weir, 1998). Not because we live under surveillance, but because we inhabit individual film sets: small perfectly coherent worlds, organized by recommendation systems and recommendation algorithms that prioritize what best retains our digital attention. This mechanism is no accident: it is part of an attention economy rooted in digital capitalism, where time spent becomes value.
Truman’s world is not spectacular. It does not offer him wealth, power, or an extraordinary life. It offers something much more effective: predictability. Truman knows who he is. He knows what to expect from life and what not to. He knows what he wants and why he does not have it. Even his frustration appears with a clear, manageable reason.
He does not travel because he fears the sea. He does not leave Seahaven because he lost his father. He does not succeed because he is an ordinary person. Nothing appears as an open wound that forces him to rethink his life; everything is integrated into a chain of reasonable explanations. And there is the most important point: his world is not designed to make him happy or to punish him; it is designed to confirm him.
Something similar happens with the recommendations that organize what we see on networks, platforms, and feeds. Their goal is not to offer us new perspectives, but to hold our attention. To do so, they select familiar content according to what we have already shown retains us. They can return it to us as aspiration or as frustration; it makes no difference. The decisive thing is that everything that appears reinforces the idea we already have of who we are and how the world works.
As in Truman’s set, nothing should be left loose for too long. If something interests us, it appears more. If something worries us, variations appear. If something retains us, it returns. The point is not for everything to go well, but for everything to feel predictable, for nothing to force us to change coordinates.
Thus, ratification replaces satisfaction. What keeps things going is not a search that opens possibilities, but a constant occupation: more similar content, more variations of the same. There is always one more video, one more opinion, one more confirmation. The flow does not stop.
That is why the system does not need to hide anything. There is no need for active deception. It is enough that what is different appears less and what is familiar takes up almost all the space. We may complain or celebrate; the mechanism does not need our agreement, only that we stay inside.
Dopamine, Prediction, and Uncertainty Reduction
The question then is why all this works so well. And the answer is not, primarily, cultural, but neurobiological.
Our nervous system is not designed to maximize pleasure, but to reduce uncertainty. The brain works as a prediction system: it constructs an idea of what the world is like and what place we occupy in it, and anticipates what may happen. When what appears confirms that idea—even if it is not especially gratifying—the organism can organize itself. When it contradicts or leaves it in doubt, the alert is triggered.
Dopamine too, which we often associate with pleasure, is more related to prediction than to satisfaction. It is not released mainly when something is pleasant, but when an expectation is fulfilled as anticipated. What hooks us is not so much full enjoyment as the repetition that keeps the circuit active.
That is why what is consolidated is not an intense experience, but an uninterrupted continuity. What attracts is that what comes next resembles what came before, that nothing forces us to stop, rethink, or start over.
This individual set, this continuous fit that seems to tell us who we are and how things are, this constant repetition of content that returns to us again and again what we like, what we desire, what we consider correct or important, may seem to work. But the body says otherwise.
This predictability acts as immediate relief. It reduces the uncertainty of the moment. It avoids the effort of stopping, doubting, or rethinking. It allows us to continue without interruption. But precisely because of this, it does not resolve anything at its core. Every small discomfort finds a quick distraction. Every doubt is covered with more content. The tension does not disappear: it is covered by the continuity.
The result is not fulfillment, but continuousness without closure. Everything seems to fit. And yet, what accumulates is not satisfaction, but discomfort: fatigue, restlessness, irritability, anxiety, a background tension that is never resolved because it is never faced.
By avoiding interruption, elaboration is also avoided. And this accumulation of unresolved stimulation is what the body registers as tiredness: the feeling of always being busy without anything ever really happening.
As we get used to environments where almost everything is legible and tailored to what we already know, our tolerance for what does not fit decreases. Doubt, uncertainty, disagreement, or difference begin to be experienced as interruption. Not only because the environment no longer trains us to sustain them, but because it is organized for the opposite: to offer immediate coherence, quick responses, a new fit every time something is uncomfortable.
It becomes more difficult to remain before an unanswered question. To maintain a conversation without a clear closure. To keep listening when something does not match what we expected. To accept that what we think can change or that what we believed was certain may not be. It also becomes more difficult to show ourselves without that constant coherence: to present ourselves to others not as something stable, but as something open, in process.
Algorithmic Personalization and Everyday Polarization
For a time, the illusion of the individual set can be lived as a personal solution. A world tailored to oneself reduces friction, saves energy, allows for orientation. The system seems to fulfill its promise: everything fits, everything is explained, nothing gets too out of hand.
But no matter how fine-tuned the setup is, no matter how much we want to sustain the fiction of that frictionless world, there is something that never disappears: others. They are always there. And the other, by definition, never quite fits. Nor should they. That’s where the setup always fails.
A subjectivity trained for years in frictionless environments learns a very specific expectation: that what appears must be legible, coherent, predictable. That answers arrive quickly. That reactions follow a recognizable pattern. That the other returns confirmation, not uncertainty. This expectation, which works when faced with a limitless feed of recommendations, transfers unmediated to our dealings with others.
We then begin to expect from people the same as from the algorithmic environment: that they confirm our image, react as they should, fit the framework that gives meaning to our experience. Like Truman, we raise our hand expecting traffic to stop. When it does not, we do not see it as a difference, but as a failure.
When these recommendation systems, increasingly supported by artificial intelligence models, reinforce closed paths, disagreement turns into polarization. The other then ceases to be simply different and becomes a problem. And the initial reaction is not usually rejection, but correction. The other is not expelled: they are explained. They are told how they should think, feel, or react so that the relationship works. If you fit, everything will be fine.
This form of closure does not appear as intolerance, but as reasonableness. But to reduce the other so that they fit—to invalidate what they see, correct their sensitivity, simplify their position—can only work under an impossible condition: that they stop being who they are.
The conflict becomes even harder when the other, in turn, inhabits their own setup. When they too believe they know who they are, what they want, and how things should be. At that point, there is no possible adjustment. There is a clash. The encounter ceases to be a shared space and turns into a silent dispute between worldviews. Everyone experiences themselves as the sensible one, the clear-sighted one, the one who is right.
A logic of winners and losers is then established: who has to adapt, who must yield. The bond ceases to open up and begins to close. And when that closure fails—and it always fails—aggression appears, often symbolically: ridicule, labeling, reducing the other to caricature, discrediting them in a sentence. It does not seek to understand or convince; it seeks to restore coherence. To feel once again that the world makes sense.
And when even that is not enough, comes withdrawal. Ending the conversation. Blocking. Withdrawing. Returning to an environment where nothing is uncomfortable and everything is confirmed. The harder it becomes to inhabit a shared world, the more we feel the need to retreat to a space that returns a stable image of ourselves. But this withdrawal does not solve the problem: it trains it. It makes the other even less tolerable the next time.
Leaving the Setup: Uncertainty vs. Ratification
Thus, what began as a supposed personal solution ends up producing a collective effect: the growing impossibility of living together. Not because the shared world has disappeared, but because what is shared stops organizing and becomes background noise compared to each individual’s setup.
This is the decisive point. The setup does not fail because it does not work, but because it works too well. Because it explains everything. Because it leaves no cracks. Because it demands nothing other than to fit in. And living like this—sooner or later—stops being enough. Not because a shining truth appears, but because a life fully ratified starts to dangerously resemble someone else’s life.
No human can live indefinitely to sustain someone else’s meaning. No human can indefinitely demand that others sustain theirs. The shared experience cannot be reduced to a performance. The other is not a piece of the set. Never was.
Truman—true-man, “real man”—is the perfect subject of the setup. He knows who he is, what he wants, and his place. He does not doubt. He does not overflow. He introduces no noise. He works exactly as he should. On set, each person is part of the setup and lives to sustain his world. In real life, without knowing it, Truman lives to sustain the world of millions who watch him.
Truman works for everyone. And yet, he fails at a decisive point: himself.
Truman does not leave because he knows what is outside. He leaves his setup because there is nothing left to discover within. Because a world that explains everything is a world that closes everything off. Truman understands that when all is decided beforehand—when he is not going to encounter anything or anyone unexpected, and when no one can see him beyond what they already expect—the only thing left is repetition.
A repetition without risk, without surprise, without the possibility of changing or being changed. A world in which everything fits, but nothing really happens.