I-Thou: Why AI Can't Give You Real Connection
You can have 400 contacts in your phone and feel completely alone. Most people know this from the inside. What’s harder to name is why — what specific thing the 400 contacts are failing to provide.
Aristotle had a name for it. Several names, actually, laid out in Book VIII of the Nicomachean Ethics with a precision that still cuts.
2026 Mental Health Awareness Month’s theme, “More Good Days, Together”, points at relational wellbeing as the core driver of those good days. The timing is deliberate. We’re in the middle of a genuine loneliness crisis, one that has persisted despite people having more social connections available than any previous generation. The paradox is the point. We are more connected than ever and more isolated than ever. And they’re not unrelated: one is partly causing the other.
Aristotle’s account of friendship is the most precise analysis of why I’ve found in any tradition, from any era. He wrote it roughly 2,350 years ago. It’s more useful right now than most of what was published last year.
The Quick Version
Aristotle distinguished three types of friendship (philia): utility (chreia), pleasure (hedone), and virtue (arete). Only the third is what he called “complete” friendship. The first two are real but fragile — they dissolve when the utility or the pleasure ends. Virtue friendships are different in kind: they’re grounded in who someone is, not what they provide, which means they deepen rather than wear out. The modern loneliness paradox is almost entirely explained by the fact that social media and contemporary life have multiplied utility and pleasure connections while making virtue friendships harder to form.
Philia (noun, ancient Greek): Aristotle’s term for friendship in the Nicomachean Ethics — broader than the English “friendship,” it covers any relationship of mutual goodwill. Aristotle identified three forms. Complete philia — virtue friendship — requires mutual admiration of character, mutual well-wishing for the other’s sake, and accumulated knowledge of who someone actually is. It is rare, slow to form, and the only type that compounds over time.
| Type | Greek term | What holds it together | Duration | Modern equivalent |
|---|---|---|---|---|
| Utility | chreia | Mutual usefulness | Ends when the use ends | Work friendships, LinkedIn contacts, neighbors you get along with |
| Pleasure | hedone | Enjoyment of shared activity | Ends when the fun or chemistry fades | Gym buddies, social media mutuals, drinking friends |
| Virtue | arete | Genuine admiration for who someone is | Deepens over time | Complete friendship (rare) |
Aristotle’s own framing: utility friendships “are those in which people wish good to each other on account of utility, not in themselves.” Pleasure friendships exist “because of pleasantness.” Both types involve real affection and can involve genuine warmth. He wasn’t dismissing them. He was categorizing them accurately.
The problem isn’t that utility and pleasure friendships are fake. It’s when they’re the only kind you have.
Work friends, gym friends, class friends, the neighbors you get along with — these often feel warmer than they actually are. You share regular contact, real humor, genuine concern when someone’s having a rough week. From inside the friendship, and from outside, it looks like the real thing.
Then the context ends. You change jobs. You move. The class finishes. You exchange numbers. You mean it. The connection fades anyway.
This gets misread as personal failure — you’re bad at friendship, adult friendship is just harder, maybe you’re not someone people want around. Aristotle would say: you’ve correctly identified that the friendship mattered, and incorrectly attributed its ending to yourself. The friendship was real. It was also utility-based. The shared context was doing most of the structural work, and when the context disappeared, there was less underneath than you felt.
Both things are true at once. That’s the point.
Pleasure friendships are trickier, because they can feel richer and more personal than utility ones. These are built around shared sensibility — people you enjoy being around specifically, for who they are in the room with you.
They’re real. But they’re still contingent — on the shared activity continuing, on the chemistry persisting, on both people continuing to enjoy the same things. When someone’s tastes shift, or one person grows in a direction the other doesn’t follow, the friendship frequently doesn’t survive the transition.
Social media has done something structurally damaging here. Instagram follows, Twitter mutuals, TikTok communities — these are nearly all pleasure connections. You follow someone because you enjoy their content. They follow you back. There’s a real sense of connection, of shared sensibility, of being known by someone.
But the Harvard Study of Adult Development — which has been tracking human wellbeing for over 85 years — is unambiguous: relationship quality is the single strongest predictor of health and happiness in later life. Not quantity. Not how many people consider you a friend. The quality of your closest relationships. Whether they involve genuine support, honest communication, and mutual accountability.
You can have a thousand pleasure connections and no complete friendships. From the outside, your social life looks rich. From the inside, you feel alone. This isn’t contradiction — it’s Aristotle’s taxonomy, playing out exactly as described.
Aristotle was specific about what made virtue friendship complete rather than partial. It’s not just more intense pleasure or more durable utility. It’s a different structure entirely.
Virtue friendship requires that you genuinely admire who someone is — their character, their way of being in the world. Not what they do for you. Not the fun you have. The specific person. And that this admiration is mutual. Both people see something in the other worth admiring. Both wish good to the other for the other’s sake — not as a side effect of what they receive.
This is why virtue friendships are rare. They require real knowledge of someone’s character, which takes time, requires honesty, and usually means having seen each other in difficult situations — not just pleasant ones. You can’t form a virtue friendship in a weekend or through a curated content feed. You can’t fake your way into one. It requires accumulated experience of who someone actually is when things aren’t easy.
It’s also why they deepen rather than wear out. A utility friendship depreciates — the shared context erodes, the usefulness eventually ends. A virtue friendship compounds. Because you’re genuinely interested in the other person’s flourishing, their growth is interesting to you. Their becoming different over time is part of what you value, not a threat to what you had. Ten years of that creates something qualitatively different from where it started.
This is the specific claim worth sitting with: Aristotle believed virtue friendships were the only kind that actually got better with time, rather than simply persisting.
Utility friendships require maintenance. They need ongoing usefulness to sustain. Pleasure friendships require renewal — the pleasure has to keep being pleasurable. Virtue friendships grow, because the virtuous activity that grounds them is itself developing. Two people who genuinely care about each other’s moral and practical flourishing will have a richer relationship at ten years than at one, almost by definition. There’s no equivalent dynamic in the other two types.
Epicurus made the related argument that friendship — specifically the kind built on character and equality — was the single greatest good wisdom could provide. Aristotle and Epicurus disagreed on quite a lot, but they converged here: the kind of friendship that produces a good life isn’t the casual, convenient kind. Both understood that it required structural commitment. Epicurus literally built a community around it.
The loneliness research on quality versus quantity of social connections is essentially Aristotle’s taxonomy with empirical backing. Quantity of connections predicts very little. Quality of close relationships predicts nearly everything. These findings keep replicating across different populations, different methods, different decades. Aristotle named the mechanism before the data existed.
Social media didn’t create the loneliness problem. But it made one specific thing worse: it created a huge volume of low-cost, high-stimulation pleasure connections that provide just enough sense of social contact to crowd out the time and energy that virtue friendship requires.
Virtue friendship is expensive. It requires showing up when it’s inconvenient. Having honest conversations that create friction. Being genuinely interested in someone’s actual inner life, not their highlights. Admitting when you’re struggling. These are high-friction activities.
Scrolling through content from 400 people you’re loosely connected to is low-friction. It produces a faint warmth — you know things about people’s lives, they know things about yours. It’s not nothing. But it uses the same social bandwidth that virtue friendship draws from, without delivering what actually matters.
The solitude vs. loneliness distinction is useful here. The loneliness that persists in crowds — at parties, on platforms, in open offices full of people you somewhat know — is the Aristotelian problem specifically. Utility and pleasure connections provide company. They don’t provide the felt sense of being genuinely known and genuinely valued. That experience only comes from the third type, and most people have very few of those relationships, if any.
Aristotle didn’t write a how-to guide. But the structure of what he described implies several things.
Virtue friendships require honest self-disclosure. You can’t admire who someone actually is if they only show you a performance. You can’t be genuinely cared for if the person caring for you only knows your curated version. This means being willing to be actually known — not performing vulnerability as a social strategy, but allowing people to see where you’re uncertain, struggling, or in the middle of something unresolved.
They require shared experience of difficulty. The clearest evidence that a connection might be virtue-based rather than pleasure-based: how does it hold when something hard happens? Utility friendships often become strained or unavailable when you’re going through something that doesn’t serve the mutual usefulness. Pleasure friendships frequently feel awkward when you’re not in a state to be pleasant. Virtue friendship is specifically what holds — or even deepens — in those moments.
Accountability is the third piece, and it’s the one most friendships quietly skip. Aristotle thought virtue friendship involved genuine attention to the other person’s character development — not just their happiness or their practical needs. A friend who only ever affirms you, who never creates useful friction, who can’t say something uncomfortable because the relationship’s pleasantness is too precious — that’s not a complete friendship. It’s a pleasure friendship with better branding.
Look at your closest friendships. For each one:
Not as a way to discard friendships that don’t pass every question. Utility and pleasure friendships are worth having and worth tending. But as a way to see clearly which type you’re actually dealing with, and to notice whether you have any complete ones at all.
Virtue friendships don’t form in optimized social settings. They form in slow, repeated, low-stakes contact — cooking with someone, working on a project together, a regular walk with no particular agenda. Aristotle’s concept of practical wisdom — phronesis — applies here: there’s no algorithm for it. You develop the capacity through practice, in actual situations, over time.
If you want more virtue friendship and have very little, the honest answer is that it probably requires creating structure for sustained contact with someone you already sense might have the potential. Not a networking coffee. Something more like a standing dinner — a standing commitment, low pressure, repeated over months.
Some people lack virtue friendship not because they haven’t found the right people, but because persistent patterns make it very difficult to form. Social anxiety that makes honest self-disclosure feel dangerous. Attachment styles that produce either clinging or distance before anything can develop. Histories that make extending trust genuinely difficult.
These aren’t philosophical problems. Aristotle’s categories are useful for understanding what kind of connection you’re missing. They don’t explain why forming it feels impossible, and they don’t address it. For that, a therapist is the more appropriate tool.
Some loneliness is also appropriate and unavoidable — the result of loss, transition, or the genuine scarcity of compatible people in a given place at a given time. Philosophy helps you think about it more clearly. It doesn’t dissolve the ache.
“More Good Days, Together” is pointing at something real. The together is the whole thing. But Aristotle’s caution stands: not all togetherness is equal. Five hundred connections can leave you exactly as alone as none, if none of them are the right kind.
The right kind is rare, takes time, and requires a degree of honesty that most people are at least a little afraid of. But it’s also the only kind that runs in the right direction — that gets richer the longer it runs, that holds under difficulty rather than dissolving when the context disappears, that provides the genuine felt sense of being known.
That experience turns out to be most of what a good day actually requires.
If persistent loneliness or difficulty forming close relationships is significantly affecting your daily life, please consider speaking with a therapist. Aristotle’s categories are a useful frame — they’re not a clinical intervention.