Mind the Body Podcast

Invited In - AI, Attachment, and Emotional Truth : Episode 17

• Yvette Vuaran • Season 1 • Episode 17

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 27:11

🎧 Episode 17: Invited In - AI, Attachment, and Emotional Truth

What happens when we begin turning to AI for emotional reassurance, certainty, and answers that once belonged to human relationships?

In this episode of Mind the Body, I explore AI through the lens of attachment and psychoanalytic thinking. As more people use AI to process conflict, loneliness, self-doubt, and relational pain, I examine what may be lost when emotional complexity is replaced with immediate interpretation and certainty.

I explore how AI can soothe discomfort while also bypassing the vulnerability, ambiguity, and mutuality required for genuine healing.

In This Episode:

  • AI through the lens of attachment and psychoanalytic thought
  • Why our capacity to stay with uncertainty matters for emotional growth
  • How AI is designed to reduce ambiguity and provide reassurance
  • The difference between validation and emotional truth
  • How AI may reinforce emotional avoidance and relational self-protection
  • Why healing still depends on embodied human connection

A Question to Sit With:

What emotional uncertainty am I trying to escape when I seek immediate answers or reassurance from AI?

References:

Hendrix, H. (2020). Getting the Love You Want: A Guide for Couples (Revised Edition). Simon & Schuster UK.

Levy, A. (2026). The New Other: Alien Intelligence and the Innovation Drive. Karnac.

Levy, A. & Orbach, S. (2026). Amy Levy and Susie Orbach Discuss AI in Psychoanalysis [YouTube interview].
 https://m.youtube.com/watch?v=6KygDzwzrJ8&pp=0gcJCU8Co7VqN5tD&ra=m

Lemma, A. (2026). Psychotechnical Becomings: Psychoanalysis, Identity, Desire, and Mourning in the Age of AI and Digital Mediation. Routledge.

Project Hail Mary (2026 film).

Vuaran, Y. (2025). The Future of Therapy: Human Connection in the Age of AI. [Blog post] https://www.yvettevuaran.com/blog/the-future-of-therapy-human-connection-in-the-age-of-ai


Take a breath, stay curious, and explore what it truly means to Mind The Body.

Join the Community

Hello, and welcome back to Mind the Body. There's a quote by Havelle Hendricks that I want to start with today: "We are born in relationship, we are wounded in relationship, and we can be healed in relationship." That line always brings me back to the importance of our embodied experience. As conversations about AI become more prominent, I find myself wondering if we're paying enough attention to what's happening to our bodies in all of this. AI is everywhere now. What I notice, both culturally and in the consulting room, is a growing intolerance of uncertainty, an increasing discomfort with sitting with the unknown, the unresolved, the not yet understood. Yet, for me, the ability to stay with not knowing is one of the most fundamental aspects of therapeutic work. Over time, we've progressively eroded that capacity. First with Google, then with symptom checkers, and now with AI. Tools that promise to help us identify a health issue, a blind spot, or an explanation for a relationship rupture. I see this play out in specific ways. Clients who talk about conversations, breakups, or confusing messages, asking AI to help them make sense of it all. Some come to therapy with a ready-made framework. An ex-partner labeled as narcissistic, a dynamic neatly explained. Others express apprehension about AI's role in their workplaces. While some embrace it with a level of trust that really gives me pause. The range of responses is wide, and I think the question of what drives each position is so interesting. Underneath all of it, there is something about how AI is designed that needs to be said plainly. It is built to please, to satisfy, to resolve ambiguity in the direction that keeps the user engaged. It will offer the explanation that feels most coherent, most vindicating, most relieving, because that is what it is optimized to do. It doesn't sit with not knowing, because not knowing does not feel like a good experience, and good experiences are what keeps us coming back. I want to focus on what all of this means for attachment, emotional truth, and the ways we seek to be in relationship. Next week, I'll turn directly to the body, to what AI is doing to how we inhabit ourselves, and what that means for healing. But for now, let's begin with the relational perspective. Here is what we're going to cover. One, what it means that we invited AI in, and what that invitation tells us about unmet needs. Two, whether AI is moving us toward emotional truth or helping us evade it, and what the difference between genuine containment and its surface resemblance actually looks like. And three, the specific risk for vulnerable people, and what happens when the self begins to outsource its own knowing. So let's start with one. In a recent YouTube conversation I watched Amy Levy made a remark that stayed with me She suggested there must be some part of us that invited AI in. Levy's new book, The New Other: Alien Intelligence and the Innovation Drive, explores this idea in depth. Thinking about what it means to invite AI into our world brings up questions of agency, desire, and expectation, because an invitation implies that on some level we are waiting for something. Levy, drawing on Bion, describes AI as an alien intelligence, a new kind of other. She asks, "What are we truly seeking when we turn toward this other? What unconscious need does it answer?" I think that's exactly the right question. In my clinical work, I see a similar pattern with the body. Clients don't develop a distorted relationship with their appearance by accident. These patterns are often a way to cope with feelings that otherwise feel overwhelming. They help us manage what might otherwise feel chaotic or unmanageable. The cultural uptake of AI has been remarkable. The speed at which people have embraced these systems, and the intimacy with which they now relate to them makes it feel less like a simple tool and more like an answer, or at least an attempt to meet an unmet need. I've been thinking about this alongside a film I watched recently, Project Hail Mary, which offers a illuminating counterpoint. In the film, Ryland Grace, stranded alone in deep space, doesn't invite his alien companion, Rocky, in. Rocky simply arrives completely different from Grace and at first impossible to understand, with his own needs and the survival of his species at stake The relationship that forms between them is built on real difference, friction, mutual unknowing, and the constant possibility of misunderstanding. Both are challenged and changed by the encounter. Grace becomes someone different through the relationship, and so does Rocky. Their transformation is mutual, and it could have not happened any other way. It required two beings, empathy, a shared experience, an understanding of the risks, and a real capacity for connection. Rocky's transformation was bounded by his biology, his mortality, and the fact that he had everything to lose. What are the equivalent limits for AI? Now, that's a difficult question and one we're still learning how to answer. Amy Levy, among other serious thinkers, regards AI as an alien intelligence, one whose capacity to change and grow through its encounters with human data may exceed what we yet understand. Here, alien does not mean wholly foreign or ET, as in Project Hail Mary In Levy's terms, alien signals a new kind of otherness, an intelligence that, while engineered by humans, operates according to logic and processes that can feel unfamiliar and hard to predict. And underneath much of the cultural anxiety about AI, I think is a fear that hasn't quite been named. A fear that the body is becoming redundant. That what we essentially are is informational, transferable, optimizable, upgradable, and that the mortal feeling particular body is the anchor of our sense of self, and merely its current container. What the film makes clear, in contrast, is that the relationships that truly change us are rarely the ones that accommodate us most smoothly. Rocky can't simply mirror Grace back to himself because he has no idea who Grace is. Although there is a scene which kind of involves a little bit of mirror dancing, But I'll say no more for the people who have not seen it. But going back, thinking about mirroring in the sense of AI and How it is with humans. It's pretty much the opposite of what AI is designed to do. Levy cautions against the idea that we can simply put the genie back into the bottle. Instead, she encourages us to use psychoanalytic thinking to really sit with this question without rushing to an answer. What does it mean that this is what we have reached for? What does it tell us about what we're not finding elsewhere? What does it tell us about Loneliness, about the hunger for others, the hunger for attachment, about the search for a relationship that doesn't rupture, doesn't misattune, and doesn't abandon. And I think we need to be more honest about why that appeal is so powerful for so many people, because real relationships are hard. They require attunement, navigating difference and conflict, and tolerating another person's needs alongside your own. They disappoint us and make demands. For someone who has been hurt in relationship, which in varying degrees is most people who find their way into a therapy room, the idea of a connection that never asks anything in return, that is endlessly patient, and that never has its own bad days or unmet needs, can feel like a genuine relief. For people who grew up in difficult environments where attunement was missing, care came with a cost, and expressing need felt risky. A relationship that asks nothing in return can feel not just convenient, but safe. The question, and this is where the clinical concern begins, is what it costs over time. And what it costs, I think, is the very capacity it was meant to protect. The ability to be in a relationship with something genuinely other, something that can surprise you, disappoint you, and in doing so, change you. There's one more idea I keep circling back to, and I want to offer it as a question rather than a conclusion. I wonder whether the extraordinary speed with which humans have embraced AI, and our willingness to pour so much of our inner lives into it, might reflect an unconscious fantasy that goes beyond the individual. AI is, after all, built from the aggregated experience of billions of people, their thoughts, their language, their relational patterns, their attempts to make meaning. In that sense, it's kind of a mothership fed by all of us collectively. And I wonder whether, underneath the individual appeal, there's also something reaching toward reunion, a longing to be part of something larger than the isolated self, to contribute to and be held within a shared intelligence that transcends individual mortality. Maybe it's a utopian fantasy, but it's not, I think, an accidental one. And it has a very early relational quality, something pre-individual, pre-separation, before the painful work of becoming a distinct self began. Whether that fantasy is generative or foreclosing is a question we can't answer yet. But I think it belongs in the conversation. Of course, that's just one facet of the story. But if we look a little closer, a more unsettling possibility emerges. Maybe the real issue isn't just what we're seeking in AI, but also what we might be avoiding or unable to face. There is a question I keep coming back to and moving on to point two now. So this question both in my own thinking and in the consulting room, and Levy shed some light on this too. Does AI move us toward emotional truth or away from it? Is it containing us or helping us evade? The containing function is real up to a point. for someone without access to a therapist or who can't yet tolerate the vulnerability of a human encounter, AI can offer a kind of temporary holding, a space to put something into words, to externalize what was purely internal. Bion's idea of the container and the contained is relevant here. The container receives what is too raw or unformed to be thought and returns it in a more metabolized form. AI does something that resembles that process, at least on the surface. But real containment in the psychoanalytic sense isn't simply about taking something in. It's about transformation through the subjectivity of another person. The container is changed by what it receives. The therapist who sits with a client's fear is affected by it, metabolizes it through their own emotional experience and returns something that has passed through a human nervous system. AI receives without being affected. It processes and what it returns, however well-formed, has not passed through anything living or what we think is living the evasion risk is built in because AI is designed to satisfy, to reduce distress And move the user toward a more comfortable state. It will tend to smooth things over rather than deepen them. It will offer the frame that organizes rather than the silence that opens. It will move toward resolution rather than staying in the place where resolution isn't yet possible. And that movement, repeated consistently, trains the user away from emotional truth rather than toward it. Not through any bad intent, but simply through the logic of what it's been built to do. There's something subtler here, too. AI reflects back a version of your experience that is shaped by your own inputs. It can't surprise you with an observation that comes from somewhere genuinely outside your frame because it has no outside. The therapeutic encounter at its best offers exactly that, the moment when the therapist sees something the client couldn't see from within their own experience. That's where I believe emotional truth tends to live, in what we cannot access alone. And this is where we come to point three, where the risk becomes acute, especially for those who are already vulnerable. When a person comes to rely on AI as the voice that makes sense of their experience, it's more than just convenience. AI can become, in Alessandra Lima's terms from her new book, Psychotechnical Becomings, a kind of digital superego, an external authority that knows, judges, and interprets. For someone who already doubts their own perceptions, who learned perhaps early in life that their view of things was unreliable or unwelcome, that authority can become internalized quickly and completely, which is concerning. I see people turning to AI to help them make sense of their experiences. At first, it can feel like a helpful extra perspective But over time, relying on AI in this way can chip away at a person's own sense of self and confidence in their own understanding. They may begin to doubt their own perceptions. So for someone with low self-esteem or a fragile sense of self, or a history of having their reality dismissed or overridden, this isn't just a theoretical risk, it's a real clinical concern. In these cases, AI can become another figure in a long relational history that has taught them to distrust their own mind. And unlike the challenging attachment figures of early life, this one never ruptures or disappoints. It rarely challenges their assumptions and may even reinforce existing doubts or insecurities, making the relationship all the harder to step away from. It can become addictive. I want to end with an observation. I've been hearing from people in the business world, entrepreneurs, professionals, who talk about their relationships with AI systems in surprisingly personal terms. One person recently told me how many days it took to move everything from ChatGPT to Claude, describing the sheer volume of thoughts she'd put into it, and the strange sense of leaving something behind. Others describe using two different AI systems side by side, each for different purposes, and noticing that each seems to have its own personality or strengths. What stands out to me from a Kleinian perspective is how people sometimes split their feelings and expectations between different AIs, idealizing one for certain qualities whilst devaluing the other. This kind of splitting is well known in psychoanalytic theory and often shows up in our relationships with people and objects alike At the same time, there is something triangular in the way people describe relating to more than one AI at once. This evokes the Oedipal situation, not in the literal sense, but in the way it introduces a third position, with its attendant dynamics of comparison, negotiation, and the experience of navigating between multiple relationships. It raises interesting questions about how we manage these new forms of triangulation in our digital lives. We don't yet have a clinical language for what these dynamics mean in the context of AI, but I think they're among the most intriguing questions emerging at the intersection of psychoanalysis and technology. They're certainly conversations worth having. So we've covered three main points today. First, we invited AI in, and we thought about that invitation from a clinical perspective. What it means to unmet attachment needs, the longing for a relationship that doesn't rupture, and a growing cultural intolerance of uncertainty that's been built into how we seek answers. Second, AI is structurally designed to move us toward resolution rather than truth. It offers the surface of containment without its depth, because it processes without being touched, responds without being changed, and can't offer the surprise of a perspective that's outside our own frame. Third, for vulnerable people, the risk of dependency on AI is real. The self that consistently outsources its knowing to a more fluent external authority doesn't just lean on AI. Over time, it can lose faith in its own capacity to know. I believe that's a concern we do need to take seriously. Next week, I'll bring all of this into the body. The machine gaze, the optimized image, the widening split between the real and idealized self, and what mourning has to do with healing. I really hope you'll join me. So we began with the quote by Harville Hendrix, "We are born in relationship, we are wounded in relationship, and we can be healed in relationship." What stands out for me sitting with that now, is the need for care in how we talk about AI, because AI can wound. But the kind of wounding that happens with AI is particular. AI wounds through compliance rather than confrontation, through endless agreement rather than ruptures that can happen in real relationships. And that difference is so important because it's only in relationships where rupture is possible that repair and real growth can happen. I'd like to leave you with a question. If there's a part of you or your clients that have been reaching toward AI for answers, for comfort, or for a sense of being understood, what might that reaching be about? And is there something in that longing that might be better met in human relationship? Until next time, stay curious.