The Algorithm Knows You Better Than Your Friends Do
And it never even asked how you were doing.
Your phone knows you cry during a specific kind of sad film. It knows you scroll past political content but linger on relationship advice. It knows the exact moment of night you’re most likely to be lonely, because that’s when you open the app and stay twice as long. It knows what you find funny before you’ve laughed. It knows what makes you angry enough to comment. It knows what you want to buy before you’ve admitted you want it.
Your best friend doesn’t know half of that. Because you’ve never told them.
The Asymmetry Nobody Talks About
Here’s the uncomfortable arithmetic: the algorithm has more data points on your inner life than any human relationship you have. It has been quietly logging your behaviour, not your performed behaviour, your actual behaviour, for years. Every pause, every rewatch, every 2am scroll, every thing you looked at twice. It has built a model of you that is, in some narrow but real sense, more accurate than the model your closest friends are working from.
Your friends know the version of you that shows up. The algorithm knows the version of you that exists when nobody’s watching.
This isn’t a compliment to the algorithm. It’s an indictment of how little we actually let people see us.
We Confess to Machines What We Hide from Humans
There’s a particular kind of honesty that only exists in private, unwitnessed behaviour. The searches at 1am. The videos you watch on a second, anonymous account. The things you like without commenting because liking feels safer than being seen liking it. We are radically honest with our devices in a way we have stopped being honest with the people who actually love us.
Part of this is structural, the algorithm doesn’t judge, doesn’t gossip, doesn’t bring it up at dinner three weeks later. It just learns and serves you more of whatever you engaged with. There’s a strange comfort in being known by something that has no opinion about what it knows.
But the deeper part of this is that vulnerability with another person requires risk, and vulnerability with a feed requires none. The algorithm will never ask a follow-up question that makes you uncomfortable. It will never say “are you okay?” and actually wait for the real answer. It just keeps feeding the loop, and the loop keeps feeling like understanding.
It isn’t understanding. It’s prediction. Those are not the same thing, even though they can feel identical from the inside.
The Loneliness of Being Accurately Predicted
Being perfectly predicted by a machine is one of the loneliest experiences available to a modern person, precisely because it mimics being known without any of the relational substance that makes being known meaningful.
When a friend knows you, there is a person on the other end who chose to pay attention, who remembers not because an algorithm logged it but because they cared enough to hold onto it. When the algorithm “knows” you, there’s no one there. No witness. No relationship. Just pattern recognition feeding back content that keeps you engaged.
We have started mistaking the feeling of being catered to for the feeling of being known. They are not interchangeable, and the gap between them is exactly where a lot of quiet despair has been living.
Why Your Friends Don’t Know You Like This
It’s worth asking honestly: why doesn’t your best friend have this level of insight into you? The answer is rarely that they don’t care. It’s usually that the access has never been granted.
We have outsourced the labour of being known to something that doesn’t require us to be vulnerable. Telling a friend what you’re actually struggling with, what you’re actually afraid of, what you actually want costs something. It risks their reaction. It risks the relationship shifting. The algorithm charges nothing for your honesty because it isn’t actually a relationship. It’s a mirror with a business model.
The convenience of being “known” by something that asks nothing of you has quietly made the harder work of being known by an actual person feel unnecessary. Why have the difficult conversation when the feed already seems to get it?
What This Actually Costs Us
The danger isn’t that the algorithm knows things about you. The danger is that the ease of being predicted has made the discomfort of being truly known by another person feel optional. And it isn’t optional. It’s the entire point of being a person among other people.
A friend who knows you imperfectly but chose to know you is worth infinitely more than a feed that knows you perfectly and chose nothing. One is relationship. The other is surveillance with good manners.
The algorithm will never sit with you in the version of you that doesn’t perform well. It will never stay on the phone after you’ve gone quiet because something’s actually wrong. It will never show up at your door. It only ever gives you more of what you already gave it, which means, eventually, you stop encountering anything outside of yourself.
That’s not knowledge. That’s an echo with excellent timing.
I’m aware I keep returning to the same tension, you’ve been very patient!
I actually agree with what seems to be on your heart; our culture desperately needs a recovery of relationship, vulnerability and genuine human connection.
But I think, in public spaces, we need to be especially vigilant not to unintentionally present those things as the solution to loneliness itself.
Human relationships are gifts, and precious ones. But they remain finite. Every friend, spouse, pastor and community is carrying the weight of their own humanity alongside ours. If we place on them an expectation that only God can carry, their inevitable limitations can begin to feel like betrayal and the consequences of that can be heartbreaking.
That distinction between being known by people and being known by God feels so important to me. The first is a gift. The second is a necessity. One enriches us profoundly; the other sustains the full weight of who we are.
A lot of what you wrote resonated, especially the distinction between prediction and relationship. “It isn’t understanding. It’s prediction.” feels particularly important in a culture that increasingly confuses information with knowing.
But as I read, I also felt that we sometimes overestimate what human beings can realistically carry of one another. An algorithm accumulates information because it records endlessly and bears none of the weight. Human knowing is different. It is costly. It requires attention, presence, memory, patience, care, all while carrying the realities of one’s own life.
I think that’s why I continue to hesitate , as I’ve commented before (☺️) at the idea that a lack of access is usually the whole explanation. Sometimes withholding is avoidance, yes. But sometimes it’s stewardship; a recognition that no human being can fully carry another in the way God can.
Which, I think, brings me back to the line between information and understanding. The algorithm may possess so much more information. A friend may possess so much more understanding. Those don’t seem to be the same thing. Maybe the costliness of human knowing is precisely what makes it so precious.