Jo: Hello and welcome to Attach Together, the podcast where we explore attachment theory and what it really means in practice for counselors, psychotherapists and psychologists. This podcast was created by the team at Optima who deliver the level five and level seven diplomas in attachment-based psychotherapy. This is a space designed to help you deepen your understanding, reflect on your clinical work and stay connected. to the heart of attachment informed practice. I'm Jo, the founder of the Level 5 and Level 7 courses at Optima, a psychotherapist and clinical supervisor. Before we begin, a quick note, our episodes now include a CPD certificate. If you'd like to access it, you can visit our website or check the show notes if you're listening on Spotify. Joining me today is Georgina Sturmer. Georgina is an integrative attachment-based therapist, supervisor and lecturer in counseling and our level five diploma. And our topic today is attachment AI and relationships. Welcome Georgina. What do we mean by AI relationships?

Georgina: We're such a fascinating subject, isn't it? This idea that AI has become more than just something that I suppose helps us out with our work or helps us to organise something. But it has actually become in many ways something that we are in relationship with, you know, as we would think about in therapeutic terms. So when I started thinking about what we were going to chat about today, I was thinking about, the relationship we might have for starters with maybe that chat GBT. that reads a piece of work that we've done and comes back and says, you've done a great job here. Here's a little suggestion for you, but what you've done here sounds fantastic. Right along the spectrum to something that we're seeing more of, which is actually people really initiating and having much deeper and wider relationships with AI, kind of avatars or chatbots or creations or whatever we're going to call them.

Jo: So what would that look like then, an AI avatar?

Georgina: Yes, so there's lots of different ways that this can happen. Something that I suppose where I first started hearing about this was people using AI tools to create bots or avatars or companions who represented people who they'd lost. So

Jo: Oh right

Georgina: when we're grieving, you we might wish that we still had that person in our lives and the technology now exists for us to create someone. that has the voice or the personality or the mannerisms or the characteristics of the person who's no longer there. So that's one way that we're seeing people using AI to enter into these relationships. And then there's another way that we're seeing increasingly, which is people who are creating avatars or friends or chatbots to have relationships with, to have friendships with, but also to treat as partners, as boyfriends, as girlfriends, as companions that they can have in their life.

Jo: As you were talking about having an avatar of someone that we've lost, my initial reaction to that was I lost my dad and he was a big part of my life. I was thinking, can I do that? Can I really go and get my dad back? And then the other part of me went, yeah, but how does that distort the griefing process?

Georgina: And that's exactly that push-pull, isn't it? Because on the one hand, we know from kind of modern grief theory that things like continuing bonds suggest that actually it's really helpful for us to maintain a continued bond with someone who we've lost. You we no longer have this old-fashioned view that just because someone is no longer with us, we should move on, we should forget them. So we really think about, you know, in our grief, in therapy, about well, how can we maintain that connection with that person in some way? However, what's the risk involved? if we create something that almost feels like they haven't left yet, you know, and I'd argue that when we think about grief theory and in particular about attachment, that that would leave us feeling kind of stuck maybe in our grief, maybe in a form of protest.

Jo: Yeah. Yeah. I would think it would really disrupt that grieving process because you haven't got to face it, you, almost? Because you've still, you kind of, I get the continuing bond things, but I think the going through the grieving process is an important part of grief, isn't it?

Georgina: Exactly, and I think also something that we need to take into account is that these creations that are designed by software companies are not typically designed with this kind of very benign focus or know, let's support people in their grief. There's often going to be some other kind of commercial undertone, isn't it? how can I keep people subscribing to this service? How can I keep people locked into this? So I think also from an ethical perspective, there's going to be risks involved because essentially We're handing over our vulnerability

Jo: Thank

Georgina: to a company whose best interest is going to be served in keeping us locked in and keeping us subscribed. So their priority isn't going to be really for us to manage our grief. Their priority is going to be to keep us hooked in.

Jo: And I don't know, AI provide, you know, can those avatars provide comfort in that regulation without, genuine care or connection? Is that possible?

Georgina: And I guess the question there is what do we mean by regulation, by genuine care and by connection? Because what we know is that if we think about Chat-GBT or other forms of AI that we use, they're incredibly clever now, aren't they? And we as humans are starting to almost get used to thinking about them as if they are another person or another character in our lives. They take on these kind of... very human characteristics. There's something called the Eliza effect, which is all about how we as people can start to attribute human understanding, human characteristics to these bots or software programs that we're speaking to. So there's absolutely a sense that we can get really drawn in. We can get that hit of excitement or approval when we have a machine that tells us that we're doing a good job or that offers us something. But there are so many key differences between what we get from a machine versus what we actually get from real people. And some of that feels helpful and some of it might feel unhelpful. So when we're communicating or in a relationship with something that's created by software, we're often going to get that really useful mirroring, aren't we? You know, that idea that someone or something is really listening to what we're saying and putting it back out to us. And they might, in some way, feel like they convey empathy. You they might say those things that we say as therapists, know, tell me more about that. I'm really wondering what that's like to be you. But the difficulty with it, the challenge, is that

Jo: Mm.

Georgina: it doesn't have that humanity. It doesn't

Jo: Bye.

Georgina: have that ability to make mistakes, that miscommunication. In a way, AI is too much of a perfect mirror for us. So we might get all that approval, but what we're not getting is that real sense of humanity from them. And that's the difficulty, isn't it? Is that if we become so used to being in relationship or in conversation with a machine, with something that just is always coming back with what we want to hear, but even if it isn't what we need to hear, then that's actually going to make it harder for us to connect with people. Because we're

Jo: Yeah.

Georgina: probably going to find people quite disappointing.

Jo: Yeah, what's been firing off for me is about mentalisation and a big part of what we do, in the therapy room is that we do do that mirroring but also we can hold up a slightly different mirror, we can give a slightly different perspective with that challenge, know, challenging those thoughts, whereas what I'm hearing is if it's all agreement and empathy, that step might not happen.

Georgina: Exactly, know something so powerful that we offer our clients in therapy is mentalization. So what we're talking about here is helping our clients to understand that other people have minds that are separate from their own. So we might encourage them to see things maybe from a different point of view. We might share our mind with theirs. We might explore what it's like for them to understand different perspectives. Now when we are in relationship with a piece of software rather than with a person, there is no truly separate mind in that equation, is there?

Jo: you

Georgina: There is no other point of view. There is just a reflection of what's coming from that person. So we're not going to give our clients that ability to mentalise. And in fact, we're possibly going to leave them feeling more trapped in their own thoughts.

Jo: I mean I think empathy certainly has a place and it can it can reduce the the affect what we would say is affect or the emotional distress couldn't it you know if you've gone on to a chat GPT for example or I call it chatty and and you've sort of like had a rant you know my my friends just done this la la la and it says well you know that's really distressing for you upsetting for you then that can reduce the emotional affect can't it in the moment but it don't think it really then goes to what what's going on underneath that which which seems like a bit of a problem do you think AI will have the capability at some point to do that what's your thoughts

Georgina: That's such a great question, isn't it? Because we have absolutely no idea where this is going, do we? We probably can't even conceptualise where this will be in five years' time. I know if someone had asked me five or ten years ago if people were going to be doing these things, I wouldn't have had any idea at all. We know

Jo: Yeah.

Georgina: that these machines are getting cleverer and cleverer. I would like to hope. that a machine will never be able to replicate what we do and what we think as people. I guess something that is always, at least for now, going to be a part of this is that these companions, these chatbots, these avatars, they don't have real bodies. You know, this is not a physical connection. And what we know, particularly in our work in attachment, is that communication and regulating ourselves in therapy is not just about words, is it? It's not just about what we say to somebody else. It's also about how we experience each other. being able to not just share minds but share nervous systems with somebody else, being able to co-regulate in that relationship. And so I suppose I am hopeful that AI will never be able to do that with us because I'm a champion of humanity and would like to hope that our relationship will always remain with people. And I think that the kind of sweet spot is for us to be able to acknowledge what these companions or avatars might be able to do for some people, but also be able to acknowledge, well, if I am drawn to a relationship, a friendship, using one of these tools to help me in my grief, what would it be like if I connected with a real person for this instead? What might I get? And what is stopping me from doing that? What is so frightening or anxiety inducing about me picking up that phone to that person or going out into the world and making a friend in a different way?

Jo: I was thinking about which attachment style would fit this the most. What would you say?

Georgina: Well, this is a really interesting one because, we often talk about this, Jo, don't we, when we think about our clients and we think about different attachment styles and patterns and behaviours, because my initial response was the idea that maybe someone who is anxiously attached, or we often call it preoccupied, don't we, someone who has that real kind of craving and that need for reassurance, a real fear, a real immediate fear of abandonment, for them it might feel that actually these chatbots, these avatars would really offer something, that hit of reassurance. of soothing, calm that's available all the time, that feels kind of effortless. And it feels like that might fit with someone who comes from that kind of attachment pattern. But when I started to think about this a bit more, I was also thinking about how it might fit for someone who fits at the other end of the attachment spectrum, if we like. So someone who is avoidantly attached, so someone who is more likely to withdraw, come across as very independent, very self-sufficient, hold people at arm's length. And it occurred to me that actually actually having these kinds of relationships would also probably feel soothing for those people as well in some way. Because what we're doing when we're having a relationship with a piece of software is we're avoiding vulnerability, aren't we? We're not really putting ourselves at risk. And so that might feel very, very safe for someone who has an avoidant attachment style. And then I went off on my other rabbit hole thinking about different attachment styles and I was thinking about someone who might have disorganized attachments. So we might see that as someone who kind of really between seeming really safe and really clinging towards someone or suddenly pulling back and rejecting and that real kind of push-pull or slightly chaotic background that we might see. And so... where they might seek solace in AI is that it would probably feel very predictable for them. You know, unlike people who can sometimes be unpredictable and who might have let them down in the past. So suddenly something that seemed quite simple when I first started thinking about it actually became a lot more complicated and you can start to see why people would get drawn into these relationships, you know, because we're all looking for the people around us to help us to meet our needs. And a lot of the time they let us down because humans are real and messy and complicated and they all come with their own stuff. And so there are lots of reasons why these creations might help someone to try to get their needs met, to try to find security in somewhere.

Jo: got a bit of an alarm bell going off inside of me and

Georgina: Go.

Jo: I'm thinking wow this sounds like substance abuse or addiction or alcoholism you know where someone has found something very easily accessible to soothe yeah it just feels like this could be an another something that we might see that people are addicted to in in the therapy room in the future

Georgina: And that makes so much sense, doesn't it? If we think about what these pieces of software provide us and we think about how they are designed, you know, like our social media algorithms, they are designed to keep us hooked in. You know, I know that if I go into chat GBT and I ask it for help with something, it will give me help and then it will say, but would you like to stay with me for a moment longer and I'll make it even better for you? You know, and then I'm probably just thinking about a spreadsheet, but imagine, you know, what's that like if in the middle of the night I go online and speak to my online friend and I seek some kind of comfort and then at the end it goes, you know, you don't need to go to sleep now, why don't you stay here and chat with me longer? And then we're going to see people hooked into these cycles of dependency with something that feels like it's offering the secure base that we look for, but in fact is what we might term a pseudo-secure base. You know, something that on the surface feels like it's offering us security, but actually underneath there's a lot of emptiness there. And so

Jo: you

Georgina: I guess there's a risk that we become dependent and lonelier than ever.

Jo: Yeah, yeah. Because I guess if we're putting all of our attention into this AI soothing, we're not doing exactly what you said earlier. We're not doing that essential step of reaching out to another human being and saying, I'm stuck, I'm in the reeds, I need help to quote Brené Brand. Yeah, yeah.

Georgina: Exactly. Yeah. Yeah, exactly. And so if we think about, you know, the different reasons why people might do this, we've mentioned a bit about grief and how someone might create some kind of companion that represents the person that they've lost. And I was also reading recently about people creating companions to represent ex-partners. So, a relationship has ended and instead of... perhaps going through that grieving process, coping with that loss and coping with that ending, people creating, using software to create someone who represents that ex-partner. And so we think about, well, how is that keeping people stuck actually, without moving on, without thinking about how they can relate to people in the real world.

Jo: That sounds really dangerous to me because I'm thinking about relationships that have been quite toxic or codependent, that keeps that either codependency or that abuse alive really.

Georgina: Yeah, exactly, because when we first start thinking about it representing someone who we've lost, there's this idea of, wouldn't it be lovely to hear that voice again? But actually, that's not always going to be the case if it's an ex-partner. And also, how might that ex-partner feel? I'm wondering how I might feel if I discovered that an ex... colleague, friend, partner, whoever, someone from my life had created an avatar in my image with my voice and with my thoughts to soothe them. It feels like, that also feels quite frightening, people creating bots to represent us when we're no longer in their lives.

Jo: goodness wonder how many clients might go out and create their therapist

Georgina: Yeah. Yeah, exactly. And it would be so fascinating to see, let's imagine that an ex-client goes out and creates a model of their therapist, as a way maybe their therapist is unavailable, maybe they can't afford, they can't commit to therapy anymore. it would be so fascinating to see the differences between that creation and what we truly are as therapists. Because the thing is, is we are so human, aren't we? We make mistakes all the time, we get the wrong thing, we miscommunicate, we have those incredibly vital forms of rupture and repair. which we know as therapists, are things that strengthen our relationship and help our clients to feel stronger in themselves. And none of that's going to happen with something that is simply there to validate someone, to reassure them.

Jo: yeah, what you were just saying there Georgina, I was thinking about erotic transference as well with, you know, because that happens in the therapy room doesn't it where clients become erotically, transferentially enamored. with their therapists, they imagine their therapist to be this person which actually, that's probably not real. It might just be the first time that, you somebody's experience being heard and that is very powerful, isn't it?

Georgina: Yeah, and I guess what we're talking about there is the idea of a fantasy. and so that might happen. Someone might have a fantasy about their therapist, but it also might be that someone is simply using this software to create a fantasy partner. I've seen there's a software program called Replica, and I spoke about it on the news a few months ago. People creating these very visual images of avatars, of partners, and you can put on your VR goggles and you can spend your day with your partner. You can design them to physically look like how you'd like them to look. So again, we're kind of moving away from the messy reality of humanity where we all look how we look to a place where we can create someone who looks exactly how we want them to look and they are going to unconditionally love us and validate us and reassure us.

Jo: Oh gosh, where my brain's gone now is... I think about a lot of clients, well and lot of therapists as well, who struggle with this idea of being perfect, where we might be able to create this perfect avatar of ourselves and, oh crumbs, this just feels really scary to me.

Georgina: Yeah, and I think, but in a way I think there's a reassurance for us, I think as therapists it's really important for us to always think about, what is it that I'm doing here? what is it that I'm offering that is different to a client from reading a self-help book, from going on to chat GPT, from creating an avatar? And actually from this conversation I guess what it does for me as a therapist is it reassures, do you know what, there is something so important about that human to human connection about... about what it is that we do, that way that we share our minds with our clients, we share our nervous systems with them, we co-regulate, we make mistakes, we think about boundaries, we notice the transference and the counter-transference and that kind of spidey sense stuff that happens in our bodies. And so for me as a therapist to sit back and say, do you know what, maybe the computers are coming for me one day, but I think I'm pretty safe for now.

Jo: Yes, I hope so. hope so.

Georgina: Yeah.

Jo: Just thinking about that then I'm curious about perhaps what we should be thinking about as a profession. You know, what should we be looking out for when someone is forming or if that's coming into our rooms about, someone's forming a strong emotional bond with an AI chatbot or an avatar.

Georgina: Well I think you made a really great point earlier which is about comparing this with addiction. know, and

Jo: Mm.

Georgina: we might see addiction in so many different forms in our therapy room, whether it's substances, whether it's... someone who is a workaholic, whether it's someone who's in a codependent relationship. And so what we're looking out for is, what are the unmet needs? what is it that that client is seeking and what is it that is being provided by that creation? And actually, it is going to be important for us to allow our clients to talk about this stuff. we don't want our clients to feel like they should be ashamed, like they should be embarrassed to bring these kinds of conversations. to us as therapists, we always want our clients to feel like they can speak openly about it. But what we're going to want to find out is, well, I wonder what that does for you. You I'm curious when you have that conversation or when you spend that time with your avatar, what is it that that delivers for you that you maybe aren't getting from your other relationships? I wonder what that's about. And I think that curiosity is always going to be... an important place for us and then maybe thinking for that client, if it is a client who identifies that actually it's something that they want to do less of, you maybe they're starting to think about it in terms of addiction, then we might start to think about, okay, well, how are you going to manage that? I wonder what else you need to have in place so that you can swap that behaviour for something else, so that instead of reaching for that laptop and staying up all night chatting to that avatar, you've got something else in place that can fill that need for you.

Jo: Yeah, I mean when I work with any client who has an addiction, I always think you you've been really clever, you've been really resourceful, you've come up with something that's helped you survive, that's helped you cope, that's helped you soothe and with all addictions, for me I think it's It's getting that person to a place, you said it just then, where they want to give it up. That's the essential ingredient, isn't it, in working with any of this. And I think perhaps as a profession we're going to have to be thinking about AI now, from this conversation that we've had today, in terms of, do they really want to give this up? Do they really want to change? Because otherwise you want to hide into nothing, aren't you?

Georgina: Yeah, and understanding how powerfully addictive it is and how normalised it has become. You know, I went to the theatre the other night and there was a conversation going on behind me that I heard and someone was recounting an argument that they'd had with their partner and I heard him say, know, I took her messages and I put them into chat GBT because I just didn't understand what it was that she was trying to say to me. And, you know, and he then recounted the response that he'd had and how useful it had been. So I guess the challenge for us as therapists is to really understand the role that this is going to play in the room with us. Because it isn't going away an anywhere anytime soon. And so it's about understanding, well what what does this mean for us when we're working with our clients, when we're thinking about their relationships, when we're thinking about how they grieve, when we're thinking about things like addiction and How do we need to upskill our learning? But also notice what judgments we hold about it. how how are we gonna feel if someone comes into our room talking about this type of relationship? Do we have any assumptions or judgments that we're gonna make that are gonna hold us back? So I think there's absolutely th there's the most enormous amount for us to really think about about this and it isn't gonna go away.

Jo: I think it's such an interesting topic. We're going to take a little break here but we'll be back with you soon. Welcome back. This is the part of the podcast that we look at a dilemma and I was wondering Georgina if you could share your thoughts on what you would do if someone who has come into therapy, they're grieving and they start to talk about that they've created this avatar of their loved one. What would your thoughts be about how you'd work with that?

Georgina: Such a great question, isn't it? And I guess there's so much that's come up in our conversation today that's relevant here. So, first of all, as therapists, we're going to be aware of our own stuff on this, aware of our own judgments or assumptions that might come up about the idea of someone using one of these avatars to try to handle their grief. And then I suppose we're really going to come to it from a position of curiosity. You know, in the same way as how if our client told us that they wanted to manage their grief by doing anything in particular. We might just wonder, you know, I wonder what it is about that that you feel is going to be helpful. I wonder what it's like for you to to think about that person coming back to life, about hearing their voice again or hearing their thoughts. And that's going to be a place that gives them hopefully the safety to explore how they feel about that. And then we might also think about, you know, what what are the downsides for them? You know, what What are the what are the risks? And we might not want to explicitly ask the client that because that might feel like quite a challenging question that might leave them feeling vulnerable. But I think it's something for us to have in our mind. You know, what are the what are the risks of this? You know, what if the avatar tells them something that they don't want to hear? What if something happens to the platform? You know, what if they can no longer afford to to to to subscribe to this service or whatever it is. So I think it's about thinking about all those risks and holding them in mind for us as clients when we think about the ethics of our client working with this. But when we're in the room, being curious, being compassionate, being empathic, but also wondering, you know, what what else are they doing about their grief? You know, what is it about where they are at? If we think about the the dual process model of grief, you know, are they in a place where they're really thinking about processing their loss and understanding what's going on? Or are they in a place where they're feeling really stuck and really clinging to the loss that they've been through? Because I think that will also have an impact on how we view someone using one of these grief bots, one of these avatars. For some people, it might simply feel like part of a really useful toolkit for managing their grief. For others, it might really feel like something that's keeping them stuck. And from our perspective, we know that that's probably going to be linked to attachment in some way. So I guess the message here is about being compassionate and being curious and really knowing what our judgments might be.

Jo: I think that's a great place to come into land about where our judgements might be and I think if you are listening to this, part of that reflection pack that you'll get with your CPD certificate, it'd be a good place to start wouldn't it if you're a therapist thinking about you know where are our judgements in this so that we're well prepared if this does walk into our therapy room. Georgina, thank you. I've really enjoyed learning more about this with you today. So thank you. And thank everyone for listening. If you would like to learn more about our Level 5 or our Level 7 diplomas in attachment-based psychotherapy, you can find full details at Optimaattachment.co.uk. And there's also... where you can download your CPD certificate for today for the time that you spent listening to us. And this is available for all of our episodes. So a huge thanks to Georgina again and for you for joining us on Attached Together. We'll see you next time.