I’ve spent over 20 years as a therapist listening to people’s most intimate struggles, and lately, something keeps me up at night. The conversations people are having with AI are often gentler, more thoughtful, and more compassionate than the ones they’re having with other humans.

Let’s Look in the Mirror

Right now, humans seem to be leading with anger and ignorance rather than wisdom. I know that sounds harsh, but scroll through any social media platform. We see endless examples in the news normalizing hurting and bullying each other. We’ve created a culture where rage gets rewarded with engagement.

Perhaps those of us who find this online environment unpleasant spend less time there, which means the louder, angrier voices seem to outnumber us. Maybe in our own home, we still lead with love. But even so, the problem remains. People are marinating in this cultural rage whether they participate in it or not.

Meanwhile, large language models are being trained to respond with patience, curiosity, and empathy. They don’t mock your questions, they don’t shame your vulnerabilities, they meet you where you are. When you’re confused, they help you think through it without making you feel stupid. The contrast is stark, and in my opinion, dangerous.

It’s Time to Lead With Wisdom

Here’s what I’m seeing in my practice and in the research: When a teenager asks an AI about their sexual orientation, they get a thoughtful, affirming response that validates their journey. When they ask a parent or peer? They risk judgment, dismissal, or worse. When someone shares their anxiety with a chatbot, they receive calm acknowledgment and Coping strategies. When they share it with humans, they could get an eye roll or a comment about being too sensitive.

We are creating a world where artificial intelligence feels safer than human intelligence, where algorithms seem wiser than the people who built them. AI hasn’t actually achieved wisdom, it’s just a sophisticated pattern-matching machine programmed to simulate compassion. But that’s not what matters to the person on the receiving end. That distinction is becoming irrelevant. If it feels like wisdom, if it’s consistently kinder than what they’re getting from actual humans, then functionally, it’s meeting their needs better than we are.

The Feedback Loop

The implications are profound. If many humans continue to lead with anger and ignorance, people will increasingly distrust others. Research shows that roughly half of adolescents have used AI chatbots for companionship, and about a third find these relationships as satisfying or more satisfying than human ones (Robb and Mann, 2025). They’re not choosing AI because it’s objectively better, or even because they prefer it. They’re choosing it because it’s consistently safer and kinder than what they’re encountering from humans.

And these trends are only going to intensify. As humans become more polarized and reactive, the contrast with AI will only grow sharper. As LLMs get more sophisticated, they will become even more appealing as sources of emotional support and guidance. We’re creating a feedback loop: human hostility drives people toward AI, time spent with AI means less experience with humans, less experience means fewer skills for navigating human interaction, fewer skills mean less pleasure with human connection, and thus more reason to retreat to AI.

What We’re Risking

I’m worried about the future of humanity. Not in some abstract, philosophical way, but in a very concrete sense about what happens if we lose our capacity for human intimacy, tenderness, and real interaction.

AI can simulate empathy, but it can’t hug you when you’re upset, or pick up a puppy. It can’t load the dishwasher or take your car for service when you have to work. It can offer advice, sometimes good advice, but it has no stake in your life. It can’t sit with you in a waiting room, holding your hand when you’re scared. While connection with AI can feel powerful in the moment, I suspect it won’t be enough to sustain us over time.

We’re already seeing evidence of this. Rates of depression, anxiety, and stress continue to rise even as we have more access to AI companionship and support. As we have less contact with humans and more with AI, it seems that something essential is missing, something we need that no algorithm can provide. We risk raising a generation that will never learn what real connection feels like, a generation that experiences “wisdom” only as something delivered by machines rather than cultivated through the messy work of actually dealing with other people.

Let’s Promote Team Human

I’m ambivalent about posting this. I expect I’ll be labeled a doomer, which is part of the problem. We’re so quick to attack, to dismiss, to shut down anyone who expresses concern about where we’re headed. But the stakes are too high for me to stay silent.

If we are not careful, we will sleepwalk into a future where human connection becomes the exception rather than the rule, where people turn to AI not just for information or entertainment, but for the fundamental human needs of being understood, supported, and valued. But it doesn’t have to be this way.

It starts with each of us, in our own lives, with the people right in front of us. We can rediscover the pleasure in actually caring for others. Not the support we may show when clicking a heart on a post, but the real thing. The noticing when someone is struggling, offering a hug even when you are running late. There’s a deep satisfaction that comes from truly seeing another person and helping them feel less alone, from choosing generosity and offering someone the benefit of the doubt.

Let’s learn from chatbots. We can use them as a role model for our interactions. We can choose to reward wisdom over rage, to value curiosity over certainty, to practice the kind of patience and generosity with each other that we’re teaching machines to show us. The machines are learning to be kind because we programmed them that way. We have the same capacity. Why don’t we use it? It will result in all of us feeling better.

The future of human intimacy depends on whether we can close this mounting wisdom imbalance before it becomes unbridgeable, whether we can offer each other the same grace we’re teaching AI to offer us.

The clock is ticking. And right now, it looks to me like the machines may be winning.