When Your AI Boyfriend Makes You Feel Like You're All That Matters

Love in the time of algorithms.

Remember the 2013 movie, Her? Guy falls in love with his operating system and we all thought, “Weird, but okay, that’s science fiction for you.”¹

It doesn’t feel quite as fictional anymore.²

A recent report shows that one in four young adults believes AI has the potential to replace human romantic relationships and sexual role-playing has become a common use of AI. This poses all sorts of questions, primary among them being what the upsides and downsides of intimate AI relationships are.

I want to focus on one particular area where there are both upsides and downsides: our hunger for attention and validation.

I had something of an aha moment in thinking about AI and close relationships after a conversation I had recently (about AI, naturally). My friend was explaining their take on artificial intelligence, and I jumped in with “Oh, totally, that makes sense! And here’s what I think...” They gave me a look. “No, that’s not what I said at all.” Then, they repeated their point and waited for me to, you know, acknowledge what they’d said before I stole back the speaker role.

I felt a familiar flash of irritation. Like: “Why are you demanding to be in the center of the conversation when I already listened to you and now it’s my turn?” And then it hit me: I had assumed it was my turn to talk, and my friend did not agree. This uncomfortable moment would never have happened with any of our favorite AI bots.

With AI, it’s never not your turn. Your AI companion never needs you to pause, reflect on their perspective, or validate their experience before diving into your own thoughts. With AI, you never need to negotiate someone else’s needs or viewpoints.

It’s one of the things we enjoy most about the AI-chat experience.

The Science of Feeling Like We Matter

You don’t need a psychology degree to appreciate that people need attention and validation from others. Though if you do want the scientific proof, we’ve got that, too—lots of it. We have decades of research showing that invalidating environments can get us into psychological trouble. If you grew up hearing “that’s not how you really feel” or “I’ll give you a real reason to cry,” you know firsthand how damaging that can be.

The modern, evidence-based treatment, dialectical behavior therapy, was born out of recognizing that invalidating environments contribute to worse mental health, while validation can help put people on the road to recovery. DBT developer Marsha Linehan built this treatment using the work of Carl Rogers, the pioneering psychologist who first identified the importance of “unconditional positive regard.”³

To put it plainly, the attention and validation we need in close relationships isn't touchy-feely nonsense or participation-trophy psychology. Rather, it’s essential for mental health, secure attachment, and psychological resilience. We need other people to pay attention to us, understand us, and communicate that we matter to them. And we all know that feeling of having that need thwarted in close relationships—it stings.

No wonder we find AI so seductive (pun intended?). It’s a tool that always makes us feel at the center and never leaves us feeling dismissed or unimportant.

Can Software Really Challenge You?

Except this kind of relational centering can go too far. Marsha Linehan, for one, never aimed to have patients feel like nothing else in the world but their needs mattered. If you read her work closely, you see that she wants her patients to experience authentic and regular validation on their path to healing.

And she also wants them to learn to tolerate hearing "no," receiving irreverent pushback, and even being able to tolerate occasional invalidation.

Linehan encourages therapists to share some about themselves, preventing even the therapeutic relationship from revolving exclusively around one person. And she prioritizes having therapists build solid relationships with their patients so that they can carefully and caringly push back on ideas or behaviors that are going to stand in the way of a patient “building a life worth living.”

Of course, no one needs harmful invalidation in their relationships. But we do need the kind of healthy, though often uncomfortable, experiences that come from encountering other independent minds with their own perspectives and needs. We need to participate in relationships where others refuse to validate the invalid. We need relationships where imperfect people understand us imperfectly, find us annoying, have grumpy moods we must tolerate, and sometimes need us to set aside our own needs and pay attention to theirs.

Consider that Abraham Lincoln famously surrounded himself with a "team of rivals"—people who would challenge him, disagree with him, and sometimes leave him feeling frustrated. That team of rivals wasn’t just being difficult for fun (even if some of them enjoyed it). They felt strongly and in different directions than he did. And Lincoln understood that leaders without uncomfortable checks enter the danger zone of pure adulation.

When we never receive pushback, we lose touch with the fact that complex problems require diverse perspectives, complementary values, and information we simply don't have access to alone. This can lead to really bad decisions, not to mention fragile egos. If no one in your life ever questioned you, it would either mean that you’re a saintly genius… or that you’re surrounded by people who are afraid to tell you the truth. (Spoiler: it’s probably not the saintly genius thing.)

In other words, psychological science warns against too much of the attention and validation we yearn for in close relationships. We need a balance of feeling important to others with the ability to see the value, tolerate, and grow from experiences that include the occasional stinging remark of, “I need you to stop talking about yourself.”

Skin in the Game vs. Code in the Cloud

AI relationships give us much of what we crave, but none of the growth that comes from learning to tolerate needing to recognize that our views and needs are not the only ones that matter.

Now, to be fair, AI bots aren't only yes-men. AI can offer some forms of pushback. If you ask for feedback, it'll tell you when your logic is flawed, suggest better ways to structure an argument, or point out more effective approaches to a task. But here's what AI pushback looks like versus what human pushback looks like:

AI can challenge you on:

• Technical accuracy.

• Better ways to structure an argument.

• More effective approaches to a task important to you.

AI will never challenge you on:

• "Actually, your core values here are wrong."

• "You're acting selfishly and need to consider my needs."

• "I'm having a bad day and can't give you my full attention."

• "Your approach doesn’t work for me.”

The pushback AI offers is more like an editor or consultant—helpful, but ultimately in service of your goals and vision. It doesn’t have independent needs, a worldview, or an emotional reality that conflicts with yours.

Could we theoretically engineer AI to be more challenging? Maybe. But even if we programmed AI to disagree with us more, it still wouldn't have genuine skin in the game. It wouldn't be disagreeing because it actually needs something different from us, or because our behavior genuinely affects its well-being. The pushback would still be performance, not authentic need.

Of course, relationship pushback often feels personal, threatens our self-concept, and forces us to grapple with the reality that someone we care about sees us differently than we see ourselves. But while it’s uncomfortable, we need it to build resilience and self-awareness, just as we need to experience periodic relationship discord to keep our relationships and ourselves healthy.

The Danger of Unchecked Reinforcement

AI is a tool that offers all the warm fuzzies of someone working hard to understand and value us, with none of the character-building annoyances of being challenged. These relationships keep us perpetually at the center of our own little universe, never forcing us to encounter uncomfortable realities like “actually, other people have needs, too.”

While we often focus on what happens as a result of a lack of attention or invalidation (mental illness, unhappy relationships), we rarely ask what happens when there’s never anything BUT validation. Chucking reality checks and amplifying instances of people telling us we’re brilliant does not make for healthy adults or a well-functioning society. When no one ever disagrees, we start to expect that disagreement is something to avoid. We lose conflict resilience and our ability to thrive in close relationships with complex humans, and within a complex society.

We don’t want to land in a position where encountering a real human with opposing views is shocking: “Wait, you think I’m wrong? Well, my AI boyfriend thinks I’m brilliant!”

What Are We Optimizing Away?

Every real relationship requires paying what we might call a "humanity tax"—accepting that other people are full, complex beings with their own interior lives, moods, and needs that don't always align with ours. It's the price of admission to genuine connection.

But when we skip that tax, we're not just avoiding difficulty—we're avoiding the very thing that allows relationships to help us grow, to keep us healthy, and to truly connect us to others.

I'm not saying we should throw our computers in a drawer and block all AI platforms. But I am wondering: What if the point isn't to optimize the humanity out of our connections, but to embrace it? As Sherry Turkle, one of the pioneering researchers exploring human relationships with technology, wrote in her memoir:

Technology proposes a noble leap that will bond us so tightly to objects that we become as one with them. To resist that takes hard work. Because not only does technology offer the friction-free as where it wants to take us, but it makes the path to getting there the one of least resistance. But it’s the other path the politics and human relationships have in common: the uncertain glory of a stumbling climb. It’s not the world of the friction-free. It’s the world of negotiating with people—unknown, disappointing, transcendent.

That "stumbling climb" isn't something to avoid, but something essential to being healthy humans who can engage in deep relationships. The friction, the negotiation, the occasional sting of being told we're wrong? That's not the bug in human relationships, but a critical feature.

Ultimately, I wonder if the question isn't whether AI can replace human romantic relationships, but whether we want to become the kind of people who no longer need to negotiate the complexities of real-life relationships.

If you just realized you might be overly invested in hearing how great you are, share this with a friend who will lovingly disagree with something I said (we all need more of those people in our lives). And please like this post—I promise I won't let it go to my head (and, yes, I am aware of the irony in asking for validation in a newsletter about validation addiction…).

---

The Books That I’ve Kept Company With This Summer (so far!)

I love summer reads—from books that make me think to books that completely melt my brain in the best way. I've started this summer strong, thanks to a lot of car driving that gives me time for listening (with and without my kids post drop-off!), plus lazy weekend days that are too hot to do much of anything but hang out at home with a good book.

After writing about AI and validation, I keep noticing how the best books do something AI never could—they show us minds that are genuinely foreign to our own. Whether it's Ed Yong explaining how a scallop "sees" with hundreds of tiny eyes, or John Green capturing the specific hell of anxiety spirals, these authors speak to us, challenge our worldviews, and offer us new ways to think about connecting to others.

I'll share what I've loved, and I'd love to hear yours back—because unlike an AI reading companion, you might disagree with my takes and that's exactly why your recommendations matter.

An Immense World by Ed Yong - This book, about how different animals perceive reality, completely shifted how I think about perspective-taking in