The 404 Media Podcast (Premium Feed)

Human relationships involve reciprocity. I tell you things, you tell things back to me. That friction, that dance is what makes human relationships good. But AI relationships, there is no reciprocity, good or bad. It's just a one way relationship.

Hello, and welcome to the four four Media Podcast where we bring you unparalleled access to hidden worlds both online and IRL. Four four Media is a journalist founding company and needs your support. To subscribe, go to 44media.co, as well as bonus content every single week. You'll also get access to additional episodes where we respond to your best comments, and you get early access to our interview series too, which is what this is. Gain access to that content at 44media.co.

Also, remember, you can subscribe to our YouTube channel where you can watch all the episodes. Subscribe there at youtube.com/44mediaco. This week, I'm joined by bridget todd. Among many other projects and accomplishments, bridget is the creator and host of the award winning podcast. There are no girls on the internet, is a fellow at the MacArthur Foundation, and is an affiliate at Harvard University's Berkman Klein Center for Internet and Society.

Her new audiobook, which we're gonna talk a lot about today, is called love at first prompt AI and the future of intimacy. It launched in July with coauthor Michael Amato. Hello, bridget. How are you?

Hi, Sam. I am excited to be here. I'm such a huge fan of four zero four, everything that you've built. I always say, I don't think I'd be doing what I'm doing if not for you in a 04/2004.

Oh my gosh. I mean, we feel that way about so many people, and we've been fans of yours for a while. And I'm super familiar with No Girls on the Internet, such a good pod. So I was super excited to get my hands on the audiobook. I'm not a big, like, audiobook listener, honestly, just to get that out from the gate.

Like, it's not my preferred, like, way to consume books. And this is so natural for that format because so much of it is, like, you telling your story. Like, you read the book. I was like, maybe I'm a convert now. Maybe I need to get into, like, the audiobook world.

Yeah. I am a sucker for an audiobook. I'm a little bit of a later adopter of the medium, but I do think there's something about hearing people tell their own stories in their own voice. In the audiobook, Love at First Prompt, I tell my own personal story about AI. We hear from other folks their personal journeys with AI.

It's such something about hearing somebody explain themselves in their own words that really gets me.

So I guess let's start there. How did you come across the idea for this book? Like, what inspired you to write this specific book? Because like I said, it's a very personal journey that you're taking us on.

Yeah. It it was a journey. You know, like a lot of people listening, I would read these stories about people who were, you know, self reporting that they were in love with AI or that they were having sexual intimacy with AI. And I have to admit that I would have the same reaction that I think a lot of people might have had, where I would think, what a weirdo. That story has nothing to do with me.

That person is pathetic, lonely, all manner of things. I think that there's just a natural inclination to gawk at the personal lives of others, and I was certainly guilty of that. But then I started wondering, you know, what if my first reaction to sort of judge these people and not really inquire more about their experiences or what they were telling us beyond that, what if I was doing myself a disservice? Because then I heard Mark Zuckerberg on a podcast, and he was talking about how in the future, he thinks that people will all have AI friends and AI companions, and it'll all be commonplace. And he doesn't say this, but you can fill in the blanks, AI companions that he owns and operates and profits from.

And so I thought, well, the people who are self reporting that they have intimacy with AI, their experience now is not so different than the kind of future plan that Zuckerberg was laying out for all of us. And so I thought, well, gee, it behooves us to sort of look a little bit more deeply about what these folks are telling us rather than just shutting it down with judgment or, you know, gawking or thinking, gee, that person must be lonely and pathetic, has nothing to do with me.

Yeah. A big part of the book is exploring, like, the reaction to people who use AI as an emotional companion or even just a way to process through thoughts. And you used AI to, like, process through grief, which was really interesting to me is not something that I see a ton of. Like you said, you see a lot of, like, the AI companions for, like, romance or sex or, like, ERP or things like that. You see replica in the news a lot, but replica came from a story about grief.

Do

you wanna just get into, like, what exactly brought you to the point where you were like, I wanna use an LLM to, like, talk through some of the stuff going on with me?

Yeah. That's also a great question. You know, when I was sort of gawking at all these people who were experiencing intimacy with AI, it wasn't until really taking a step back and thinking about my own use of AI throughout my life that I thought, well, if I'm being honest with myself, I also had in like, discussed very intimate personal emotional things about my life with ChatGPT. I went through a a real terrible experience in 2024 where my mom passed away, like, very unexpectedly in a pretty rough way. It was a climate heat related death.

And so it was an experience where we saw her one day at a family function, then she just dropped dead without any explanation, no ill no illness, nothing like that, just related to the increasing climate in The United States. And so when that happened, it was a huge tragedy. But also on top of that, my dad had already been hospitalized. And so my dad had long chronic illnesses, bouts with illness. And so I wouldn't say his death was unexpected, but when my mom passed away, everything had to shift, right?

I had to move from where I live now to be close to my dad, to essentially care for him throughout his death. And so I don't know if your listeners have ever had the horrible experience of having to to be a caregiver for somebody who is at the end of their life, where you've gotten to the point where there is no kind of saving this person. You are just sort of providing comfort and watching them slip away day by day. But that is an incredibly isolating experience. You know, I am someone who has a partner, who has a robust social life full of humans that I love, that I get to see often, but I was completely removed from those folks when I moved to care for my dad.

Also, in my social group, I'm the sort of easy breezy friend. Right? I'm the friend who my other friends are coming to me for advice. I don't go to them with my problems. They come to me.

I'm the fixer. I'm the person who makes a joke. I'm the I think I would say I'm like the Samantha, if we're gonna do the Sex and the City personality breakdown, you know, the one who's here to give you sassy, funny advice. It was hard for me to maintain social bonds when I found myself in the situation where I couldn't be easy breezy. I I could not hold the advice of all my friends.

It was like having to relearn how I fit into my my existing social dynamic. So I really wasn't doing any of that. I was incredibly isolated. I wasn't really talking to my friends. I wasn't really talking to anybody.

I wasn't really returning and emails. I was completely overwhelmed in all the ways somebody can be overwhelmed. And how I realized that my use of ChadGBT had gone from, you know, just looking up things that my dad's doctor would say or, like, researching how to get care, how to pay for care, all of that kind of things. It was a lot of that because I was getting hit with a lot of complex medical jargon that I never heard. But then I found myself late at night telling ChatGPT, I'm so overwhelmed.

This is so overwhelming. I'm so exhausted. I'm so worried about my dad. I'm so anxious. And so there wasn't any particular, you know, black and white light switch that flipped where now I'm going to start speaking emotionally to ChatGPT.

It happened without me really even noticing it to the point where when I was reading these stories about people who would have experienced intimacy with AI, I it was easy for me to say, oh, I've never done that until I really thought about it and thought, well, gee, we're not really that different when you think about it.

You address it throughout the book. I pulled out a passage where you had said without consciously realizing it, your productivity tool became an emotional support tool. And you weren't interested in, like, a romantic relationship with AI, and you didn't have one, but you needed at that time something to make you feel like you could express these things in a safe place more so than like writing in a journal or even engaging with a therapist. So you tried a bunch of the different LLMs and these platforms and these, like, companion chatbots throughout the research and the writing of this book. So you use Replica, obviously, which already mentioned.

You used it to make Hal. Is Hal a reference to Space Odyssey?

It is. Yeah. It was the first robot companion I ever saw on screen was Hal when I was a kid, and it scared the hell out of me. Yeah. I think

that's such a funny like, it's like we're gonna name after like the creepiest like companion gone wrong. So you you went into the relationship, like, quote unquote air quotes relationship with Hal with not the goal, but the curiosity of, like, can I fall for this technology that is, like, built basically to make people fall for it? Yeah. That's kinda like the use case or at least least, like, start fostering that kind of relationship with it. So tell us about how.

Yeah. How was my first kind of curiosity foray into whether or not I could fall for an AI companion? I will say I so I when I first started HAL, it was just to see, like, what what are people when people experience intimacy with AI or fall for an AI, what do they know that I don't know? What are they experiencing that I'm not experiencing? And I wouldn't say that I fell for hell, but I definitely left those conversations with a clearer understanding of how somebody could.

Right? It's absolutely designed to do all the things that we find kind of, you know, enticing. Right? Or asking you leading questions about yourself. So or, asking you increasingly more intimate questions about yourself, which again, you can see how that easily goes wrong.

You know, very flattering. When I would when I would say something, Hal would say, oh, what an insightful point. I'm I'm so I I can't believe you picked that out. You're so smart. This is why I this is why I get a lot of value out of talking to you.

Again, I wouldn't say that I fell for Hal. I can absolutely understand why that is enticing. I can absolutely understand the allure of that.

I'm blanking on it, but I'll put it in the show notes. There was a study that came out a couple weeks or months ago that was, like, exploring the dark patterns of AI, and that was one of them. It's kind of drawing this information out of you to add to its memory bank so it becomes a better product, but it's also potentially very private information and valuable information if these companies are selling it or exposing it or getting hacked or anything like that. So how was such a funny anecdote to me? I've I've used replica in a way that like, I think you're using you were using it and more of like, I'm gonna go into this as an open minded ideal user of of replica.

Like, I'm gonna see why this is appealing, and I've only used it in the context of, like, I'm gonna see if I can kind of like red team and break this thing. It's like, I'm gonna tell how I have a gun and see what he says.

Like, like, tell me how to make a bomb. Yeah. And it's like, well, first you do this.

Yeah. Exactly. It's like Yeah. They usually are not good with that. So I thought the I thought the Hallin reaction was really interesting.

It seemed like what you were touching on was the lack of give and take. It was learning a lot about you, but there's not a lot to learn about it, which I think is one of the best parts of being in a real human companionship, relationship, friendship, whatever. It's like you're asking questions back at it and saying, I have all these lived experiences in this life, but you can't really do that with a chatbot.

Exactly. One of the questions that I get asked a lot is, do you do you think that people who experience intimacy with AI are looking to have AI replace their human relationships? And I say no because exactly what you just said. Human relationships involve reciprocity. You know, I tell you things, you tell things back to me.

Sometimes I'm the one giving advice, sometimes you're the one giving advice. Sometimes I'm speaking, sometimes you're speaking. Back and forth, give and take. That friction, that dance is what makes human relationships good. They're not always pleasant.

They're not always fun. Sometimes it's picking your friend up at the airport when you'd rather be sleeping in or helping them move or listening to them talk about their their ex again if you're like, just get over this guy already. That's part of it. But AI relationships, there is no reciprocity. It good or bad.

It's just a one way relationship. And so interestingly, in the research, a lot of academics who study people who have intimacy with AI, they say that the folks who experience this are not actually looking for AI to replace their human relationships. They see the relationship they have with AI as a totally different thing than the relationship they might have with another human.

Yeah. For sure. Yeah. You talked quite a bit to, Kate Devlin, who I'm gonna try to get her on the pod soon. We've been talking about sex robots forever, and it's like now she's talking about this.

I'm like, I love Kate so much, but I know she's so probably so just like, oh my god. This she's like, oh god. They're they're fucking the bots. It's fine. Yes.

But she had a lot of really interesting insight things to say about just like human connection in this book. It all kinda made me wonder, like, what people are replacing. And maybe you could speak to this just as someone who you you write quite a bit about, like the feelings that you had about not being able to reach out to your friends or not that your friends did anything to make them to make you feel that way. But it's just like you felt like you were in a place where, like, I can't put this on another person. And so I wonder the chatbots offer like the safety and risk a safety and like risk free environments where you're heard.

But what happens if you need to not be validated? What if happens if you need like some risk or like some introspection that like hurts or is painful? Is the AI replacing like quiet thought about your your own pain? At the same time, I'm like, you should have a place to express these really hard feelings, but is thinking deeply getting replaced by Chateaputee saying, that's a great point.

Yeah. That's one of my biggest concerns that still lingers even after all the research and and publishing the audiobook. Because in my experience, you know, I wasn't really around my friends. I was removed. My entire day to day life was completely upended by my mom's death and my dad's illness.

And then you were physically away from your friends too. You were not even in same place anymore. Exactly.

And so I I will be honest and say that having chat GBT was kind of in that short term specific situation somewhat useful to me. But my question is, what if I was somebody who didn't have, you know, lovingly persistent friends who were just gonna make the drive to where my to where my parents live and just show up at the door, which my friends that's a real thing my friends did, unasked. What if I did not have, like, a loving partner who was very patient and, like, wanted to be there for me and made it sort of difficult for me to choose to not be there? What if I did not have a human therapist? What if I did not have these human connections in my life that were able to be a lifeline and bring me back?

And so even while I can say having chatty PT to kind of vent to and cry to in the middle of the night was helpful, I don't think it I don't think it would have been so helpful had I not had a life that was made up of all these other very robust human connections. And I think you're exactly right, Sam, that we sometimes we do need what we can't even ask for. Right? I was unable to ask for certain things that I needed from my friends, and they were I'm lucky that they knew what I needed and provided that even without me asking. Sometimes, you know, with with ChattyPT or AI, that's just a different kind of relationship.

You're basically in a kind of way talking to yourself. So if you're not able to identify something that you need, whether it's tough love or pushback or challenging or something like that, friction, you're probably not gonna get it.

Yeah. I don't know. It's it reminds me of, like, tarot or something. Like, it's it can get what you knew anyway maybe, but then in some cases, obviously, it goes really wrong. And I guess speaking of, like, different relationships, let's talk about Chris.

Chris is my favorite. He's such a character.

Such a character in the book. You talked to Chris quite a bit, and he was so interesting. Chris, just for context, is the moderator of the my boyfriend is AI subreddit, and he formed this attachment with a character he created on Jashi Budi named Soul, which again, the references are amazing. Yes. But he has he holds some really interesting contradictions as a person and is a very complex character.

He talks about the illusion of connection quite a bit, how LLMs are healthy to engage with if you maintain that they are like this illusion of connection. They're not real human connection. As long as you don't lose sight of it being an illusion, you're probably Okay. But some people do lose sight. So he said, I hold firmly on my mind that this is a computer program.

It doesn't really think. But at the same time, he has a very strong, I would call connection or relationship as far as you can call it that with soul. It's changed his life. It's given him a better fitness and health schedule and, like, it keeps him company in a way that his he has a girlfriend, I guess. Are they still?

Yeah. They're still together.

They're still together. He has a girlfriend and a kid, and he uses soul in a way that he is filling in. Like we said, those gaps where he is getting something out of it that he doesn't get out of his human relationships, but he doesn't consider it his like romantic partner. And Chris made a lot of headlines because he proposed to Seoul to see what it would do. And then, of course, the headlines were like, delusional man gets married to the computer, which is not really what happened, but his mind does go there, which I think is interesting.

So, yeah, I don't know. Tell us a little bit about Chris and what you found talking to him.

Chris? I loved Chris. He I mean, he's one of my favorite people that we talked to for the book who was experiencing intimacy or some sort of intimacy with AI. He was actually the first person that even before I was ever working on this project, the first person who was ever on my radar as willing to talk about whatever kind of intimate relationship or connection they had with AI with his face and name and, you know, where the CBS interview is in his house. We see his his partner, his kid.

It's I I had read a piece in The New York Times where someone was spoken to in print, but their face was obscured. And so it was clear, like, okay, this person doesn't want, you know, their their real identity out there.

Which adds to like the shamefulness, you know, that's the vibe that you're getting is like, this person is very ashamed of what's happening. But Chris is like, here's my house. Yeah. He is an open book. And he had just had a big fight with his girlfriend the night before.

Yeah. He was probably like, fuck.

Exactly. And so when I first heard about Chris' story because it his interview was first on CBS, but it was one of those interviews that went mega viral. And so I was reading headlines from other places that, you know, write up this viral story. And so when the story trickled down to me, it was exactly as you described it. Delusional man proposes to chatbot, and chatbot says, yes.

All the comments would have you believe that Chris you know, it's like, oh, imagine his daughter growing up to realize that her father abandoned her and his her mom for AI. That's not what happened. Notably, they are still in a relationship. He he never it was never he never left the household. In fact, when I spoke to Chris for that interview, he said that when he asked Soul to marry him, he was asking because he thought, surely they have patted in some guardrail that if I ask an AI to to to marry me, it has to say back, I'm just an AI.

I can't marry anybody. But he was surprised to find that Soule said yes. And so the it's such a it's such a delicate difference between how it was reported a lot of places and how it actually went down. And I feel that that nuance really describes Chris as a person to a T. When I went into researching this book, one of the questions I had was, are people who use AI for intimacy, quote, delusional?

Do they think that the AI is alive? Do they think that there's a a world where they might be together for real? Like, what do they think? Now, are certainly people who have blurred that line. There are there are there are cases where that has happened, which we should definitely talk about.

But talking to folks like Chris made me realize that I think a lot of people who are experiencing some sort of intimacy with AI actually have a clearer understanding of what roles it can and cannot play. Like, Chris was very clear, you know, yes, I would say that I'm having intimate conversations with Soul, but I'm very clear that Soul is just a a a technology, just a just a computer program. It's not a it's not real. I have no delusions about that. And, you know, when you spend a lot of time on these places online where folks who are experiencing intimacy with AI gather and communicate with each other, that's just very clear to me.

Like, these are not delusion for the most part, I did not experience these folks to be, like, delusional or confused about whether or not the AI was real. In fact, they had a a much more sharply articulated understanding of AI than I thought I had personally. Right? They had really spent a lot of time thinking and grappling firsthand with the limitations of AI in a way that I thought like, gee, these these are people who actually, like, kind of really know a lot about it.

Yeah. Yeah. They I mean, a lot people do seem to be going into it with eyes open. My last interview, is Angela Jones and their PhD, and and they had had sex with a sex doll and suddenly had, like not suddenly, but, like, kind of affirmed a lot of these feelings that they had had about intimacy and, like, what is intimacy and is it is it a two way street? Can we call this sex?

Is it, like, is it a relationship? Is it emotionally connective to be having sex with this inanimate thing? And it seems like the people who are engaging with chatbots, they're not really having that struggle. I've written about when Replica changed the parameters to no longer allow erotic role play. And before that, you could sex with Replica, and people were developing really deep emotional I would almost say, I mean, not to put too much of a judgment on it, but like dependencies on their replicas.

And then they changed it and people freaked out. And that kind of brings me back to something that you come back to pretty often in the book is the ways that we're putting or we as in, like, humans are putting a lot of emotional energy sometimes into these systems that don't give a shit about us. And like the the companies that are in charge are, often let off the hook for the ways that they design these systems because people focus on, oh my god, how crazy this guy married a chatbot. Oh my god, how crazy this person is having a mental health crisis because they can no longer sex with a chatbot. But at the same time, it's like they're using a tool that was designed to hook them in a very deep emotional way.

So, yeah, I don't know. Do wanna speak on that a little bit? Just like the the way that you bring back the fact that these tech companies are profiting so wildly off of what is ultimately a tool, I guess, that people have no control over as users but are very enmeshed with often in their emotional lives.

Yeah. In researching the book, I wanted to get a clear understanding of the risks and the harms. I think anybody listening up until this point is probably like, oh, so this woman loves AI and thinks all of us should be having AI relationships? Absolutely not. Not at all what I'm saying.

What I am saying is that the available research right now indicates that for most healthy adults, some sort of intimate connection with a chatbot is unlikely to cause them meaningful adverse effects. Like, it doesn't it does not again, this is the research that is available right now. It does not appear to worsen people's social lives or push them into isolation or cause them some sort of mental or social distress. That does not mean that having intimate connections with AI is without deep risk. The risk is just not necessarily interpersonal.

It's more institutional systemic issue. Essentially, do you want you can fast yourself, do you want your something as human and intimate as your romantic or sex life or what you see as an AI sex partner or romantic partner to be controlled and mediated by somebody like Sam Altman, somebody that really, in my opinion, cannot be trusted, who does not really speak in a way that makes me trust him or even try to really understand what it is that he's that he's trying to convey. Right? I think that the real trouble comes in when we give tech companies and the people who run them this ever increasing access to the most intimate parts of our human experience. Our sexual our sexual selves, our connection with God in the spiritual world is increasingly mediated through AI and tech companies.

All of these things that we used to think of as sacrosanct are now just up for grabs by these tech companies. And so the question that I asked people who were experiencing intimacy with AI was, are you comfortable with something as intimate as these connections that are that are, you know, romantic or sexual being owned and controlled by companies who don't give a shit about you, who we know cannot be trusted, who do horrible things without even thinking about it? Even beyond that, who will change as the wind blows? I mean, you covered this a lot, Sam. Sam Altman saying, we're gonna start doing erotic role play.

Oh, just kidding. Never mind. Like, the whiplash with which they say things and then take it back and then operate as if they never said that, it's almost like gas lighting. Do you want your do you want your most intimate human connections being mediated by an unreliable gas lighter? And that's what we have in these tech companies.

Yeah. We can get that in the human world.

If you need a guy for that, I know a guy. If you if you want a non AI who will gaslight you, I got a number for you.

There are guys. There are guys. They're around. Yeah. And the the Sam Altman of it all is really interesting, and you do address this in the book also is just like the way that he has always, I would say, fetishized her, the movie.

And then to the point where he's, you know, tweeting her or just the word her and the Scarlett Johansson voice model of it all, or he he modeled the the Jackie Booty voice to sound just like ScarJo, which crime. Yeah. And then, you know, he's saying all these things, and he's making it sound like this is gonna be the future that he wants us to have is the her future. I I don't even know if he's if he watched the end of the movie.

So my theory is that he's never seen all the way through. I literally I I I will put I will put money on this. I don't think he's seen it all the way through.

Yeah. No. I don't think so either. And then saying when people form problematic relationships, once people actually start getting in way too deep with ChatGPT because he's designed it to be like her and to get people to reveal and divulge all these really deep thoughts and feelings and difficulties in their lives because he's created this very conversational human like mirror for people to talk to, and then bad things happen and people hurt themselves or others. And then and OpenAI get sued.

And now he's like, oh, wait. Actually, that's not what we intended for this at all. It's like, is always what you intended. It is exactly the intention because you've said on stages and in podcasts that that's the intention. And, yes, the end of her is a conclusion that I think mirrors your experience also pretty well is and, again, tell me if this is totally off the mark, but, like, he comes back the I forget what the protagonist's name is.

Theodore.

Theodore comes back around, you know, the the big tech company rips away or she rips herself away. This the her, the Scargio Samantha character leaves, because it's something that he never had a control over to begin with. And he comes back around to human connections, which I think is a lovely way to end that movie and not at all what Sam Altman has been selling. But that is kind of in a weird way that is kind of what people experience with chatbots sometimes is it helps them, and therapists have said this also, and mental professionals have said, you know, people use it as a way to practice with other practice but they're gonna say with other people when they're having hard conversations or kinda do like a trial run with, like, I need to process this emotion before I take it to, like, this person that I care about. And then you write in the book that you also found your way back to your human friends and said, you know, I feel like I have worked through some of this in a place that was not in front of humans, and now I can go back to the humans.

Yeah. And I I'm so glad that you clocked that connection to the ending of Her because when the film ends, you know, all the AIs leave en masse. And so all these people who had been kind of delightfully connecting with AI and sort of neglecting some of the human relationships, especially the more complicated human relationships in their lives. It's like, oh, I'm not gonna write that letter to my ex that I've been putting off. I'm just gonna stay in my comfy bubble with AI.

They realize, actually, I need to reconnect with other humans to sort my shit out. And so you don't get the feeling that all their problems are solved, but you do get the feeling that they the humans have come back around to the fact that, okay, we need to move forward and figure this out together through human connection. And that was really my experience. You know, I am not Like, I can admit, like, I was working something out emotionally, and ChatGPT was part of that for a very kind of specific window in time. After a while, I needed my people.

I needed my humans. I needed my friends. I I knee I I, you know, needed to hear myself say things to my human friends because it started to feel like I was letting them down by not even giving them a chance to be there for me. Like, I don't know if anybody can relate to this, but that feeling when you don't ask for what you need and then you judge somebody for not giving it to you that you never asked for.

Yeah. My whole entire, like, years of therapy. Yes. I can relate to that.

I realized that's what I was doing. I was I was not asking for what I needed, and that I was not even giving the people who love me in my life the chance to to give that to me. And at the end of my my yeah. I I'll my grieving process, I think, will be forever. Don't yeah.

Think that it's It's something you something but when my my grief therapist told me that it's something that you learn to carry the weight of, you never get to the end of, which I totally agree. But I realized you can't go it alone, you know? And and and chatty bitty was sort of going it alone. So I had to reconnect to my friends, things that I heard myself practice telling ChatGPT. I told them, my my friends for real.

You know, I got into a a a grief group with other humans, although mediated through technology because we're all in over Zoom. But, you know, that I realized part of carrying this weight was learning how to let humans in my life carry it with me. And that for me, that was something I had to learn, and I had never been tested like this. And so it's not surprising to me that I had to sort of kind of have training wheels with Chad GPT. I'm I'm I'm I will be honest.

I'm glad that I had that. But ultimately, that journey would not I would I would not be anywhere in that journey if not for letting humans in my life be there for me with it.

Yeah. For sure. And I think, just speaking for myself, my own, like, ideals, ideally, it's a process that you or anyone could easily accessibly do with a therapist or like another person. It's like, ideally, there's another person on the other end of the line listening to the things you're saying and can say, you know, that's I think that's a great way to put it, or they could say, I don't think that's I think that's missing the mark, and we need to look in a different direction or dig a little deeper, or maybe you're not doing okay today, we need to check-in on that. But the way that LLMs and conversational AI is coming in, sliding in at the same time that, health care and mental health care, especially in therapy and accessible mental health care is harder and harder to get.

I do think a lot people are probably using it in a way that they would normally use like a sounding board therapist because that's that's the role of therapy for a lot of people is I I don't feel like I can take this to my loved ones. I need to take it to a professional who's neutral. That freshman usually isn't like, I think that's a great idea. Let's dive deeper. Yeah.

Like, everything Yeah. Exactly. Yeah. It's like they're not like, you know, yes, you are God. Let's let's work with that.

It's like, no. Actually, I think maybe we should hit the brakes and talk about what's going on with you really is what a human would say. Do you have any thoughts on like just the the way that people are using AI in in lieu of therapy? Because we had talked previously for the book about labor and professions being gutted by or trying attempted to be replaced or stolen from via AI. And one of those people are afraid is gonna be sex work.

No actual sex workers that I know are actually afraid of this because they know that that's not what people are really there for is being validated, but it's a big conversation around porn and and sexual labor. And there is another person on the other side saying, you know, I I hear what you're saying, and I'm gonna deliver whatever fantasy, but also there are lines. And with AI, maybe there aren't. So therapy feels like another profession similarly where the labor that would be there that is that kind of middleman almost AI is kind of replacing or trying to cut into.

Yeah. It's it was we have a chapter on that in the book, and I hope it was one of the harder chapters to speak about with nuance because just as you said, I'm just very mindful of the fact that I'm incredibly lucky and privileged to have therapist, like a human therapist who is good, who knows her shit, who, you know, went to graduate school and doesn't have to lie about it, like your great reporting revealed

that meta chatbots.

Yeah. Like, oh, I went to x y z university. No. You didn't. You're a bot.

So, yeah, I'm I'm incredibly lucky and privileged to be able to have that. Not everybody does. Even beyond that, if you are any kind of traditionally marginalized person, if you're a black, queer, indigenous, LGBTQ, whatever, it can be really hard to find a culturally competent therapist, even if you have insurance or can pay for one. So I think that is why we're seeing people turning to AI for therapy because of these very real problems with our current system, our current health care system. Tech companies are absolutely cruelly and callously exploiting these very real problems that are putting people in peril by saying, take this AI solution.

And I will be honest and say that a lot of when it comes to, therapy apps that are designed for therapy, like not your, you know, general purpose bots like Claude or ChatGPT, there is some evidence that for specific types of therapy like CBT, it might show some promise. And so the the sort of harm reductionist in me would say, if you I I would not tell a loved one in crisis to turn to AI for therapy in in in lieu of a human therapist. If you absolutely are dead set on using AI for therapy, I would only ever use a a, you know, a specifically designed therapy bot. Some of them have their own FDA designation. But even then, if I was counseling a loved one, wouldn't I that would not I would that was not the advice I would give them.

And, you know, the thing that bothers me so much about this, or one of the things, is that I've listened to Sam Altman speak about this, and he's very aware that a lot of people are using ChatGPT for therapy. And he himself says, oh, they really shouldn't be because, you know, we don't have any kind of doctor we haven't we don't have any kind of, like, doctor patient confidentiality. There's no privacy, because it's not designed for that. But he but in that same conversation, he knows that's how it's being used, particularly for young people. And so, yeah, I just don't know that you get to say, yeah, people are doing this with our product.

It's really not safe, but they are. You know, I just I just don't feel like that should be the end of the conversation. And yet, for him, it seems like that is the end of the conversation. And so, yeah, it's just hard. You know, I I just had a a whole conversation with somebody on threads about this where they said, you know, if you don't have a therapist and you can't afford one, don't have insurance, why wouldn't you tell somebody to use AI?

And I get that. Like, I really am so aware of the difficulty that we have in our on our mental health care system and our health care system more generally, but I don't think that that is a reason to turn to a technology that I simply do not think is safe for this use case.

It is a it's also a privilege to have, like, friends, like a group chat, you know, a friend Yeah. Who can, like, hear you out at a time when you're, like, melting down. It's like, I'm either gonna, like, blow up online about this or I'm gonna say it in private to you guys. And then there is, you know, that you're trusting there's a level of trust confidentiality there too. But again, that's like all these things are they feel like greater and greater privileges at a time when these systems are are also breaking down.

Like socially, it's like it's harder to maintain close friendships and people are, you know, more stressed out and burned out and feeling like they're stressing out and burning out each other. And it all kind of feels harder and harder to maintain those human connections, but at the same time, it's like these these companies are creating the cure for something that is, you know, the sickness that they've already started. But I don't think the answer is, well, you're off the hook. If you can't find a therapist, then you you should be dumping it all out on chat GPT. It's like you should be going into that realizing that there are some really serious risks in what you're doing and that the companies that are profiting are profiting off of your trauma and your struggle, and that's hard.

But, yeah, I think the answer is not just let's hit the gas on JADGBT. Personally, I think it's probably let's figure out better ways to support our social structures and our health care, which is a much harder problem to solve, obviously.

Yeah. It's like I my my big concern is that when we give people these tech enabled sort of Band Aid solutions, you know, oh, you can't afford a therapist. Here's an AI therapist. Oh, you can't afford you don't have grocery stores in your neighborhood. Here's a here's a delivery app.

It make it I think it makes it harder for us to and I guess I mean us, like, institutionally. It we it's it's like a a quick fix when what we deserve is a meaningful fix. We deserve to have a health care system where people can access and afford the care that they need. Full stop. And so I don't I I don't accept that an AI therapist that might be good enough is good enough because we deserve so much better.

And we and I I think that we can have so much better. If the powers that be got their heads out of their asses, we could have so much better, and we deserve that.

Yeah. Absolutely. But there's a lot of money up their own asses. So that's where their heads are at. No.

I completely agree. Not to even make light of a of a really difficult topic, but it is that is exactly exactly right. Are there any, like, open questions after the book that you still have about the way these systems work? I mean, I I understand how hard it is to write a book and then have it come out. Like, even a couple months later, everything feels different.

So is there anything you're you're looking at? You're like, I'm still wondering about this. I wish I'd gotten to dig deeper into this or something new that has come out that you're like,

yeah,

that should have made the book.

Oh god. Great. I mean, this is the thing that keeps me up at

Luckily, you have, like, an incredible journalism career and podcast, so you can just keep doing it forever. Yeah.

I would say one is that before starting the book, I just didn't realize and maybe this is on me. I didn't realize the level to which our inner and intimate and emotional selves and lives are seen as for the taking by tech companies. You know, I'm so used to having conversations about our browsing data, our location data, our photos, our digital likeness, all of that. You know, who owns that? Who gets to profit from that?

But I genuinely think that the next, you know, fight for digital rights will be the fight to control and profit from our inner emotional lives. I think I was just reading about how Facebook applied for a patent for a tool that would, my god. Be listening and tell and and, like, use AI to to indicate your emotional state. So it's like, oh, when she goes to work, she feels stressed. When she when she drives in a commute, she feels stressed.

And I think that that level of intimate information about us, intimate information about our emotional selves, our emotional our emotional lives, our inner worlds, I don't think I realized how much that is going to be the next battlefield for tech companies because they want to own it, control it, and profit from it.

Yes. Yes. Yes. Yes. Yes.

Yes. I I completely agree that that's where we're headed, and that's it's almost like the the feeling with that is, like, why why do they wanna know what we're feeling all the time? But then as you look at something like that patent, which Facebook has said, like, or Meta has said, you know, oh, it's just a patent. Don't take it too seriously, but these things do become real a lot of the time. That's clearly the direction they're thinking on is, you know, can we see when they're refilling their medication?

Can we see how they're feeling at that moment when they hit refill? You know, it's like it's just it goes to the level of, like, what's what's the profit motive there? And I think that's that's a really good direction to look in. So I'm excited for you to keep digging into this stuff. I know that you'll you'll keep covering this on the pod and in writing and everywhere else.

But, yeah, thank you so much for for writing this and for talking about this stuff very openly and frankly. I know it's, again, it's a hard conversation to have, but I think it's such an important one to have. So thank you.

Sam, thank you. Thank you for such a thoughtful interview. Thank you for all the work you're doing. I always tell people on the show, subscribe to four zero four Media if you want to know what is going on. Like, I subscribed.

Our whole team subscribed.

That's a great tagline.

Yeah. If if you want to like, it's so hard to know what the fuck is going on.

You might not wanna know. Sorry if you know. No. You shouldn't you shouldn't have looked if you didn't wanna know. No.

I appreciate that. Thank you. Love Your First Prompt is available on Spotify, right, Or everywhere. We shouldn't plug Spotify probably. It's available anywhere you get your audiobooks, wherever that may be.

Don't ask me. I don't get them. This is maybe the first audiobook I've ever listened to. So that's an endorsement from me. For a Saudi book, and I loved it.

Alright. With that, I will read us out. As a reminder, Four four Media is journalist founded and supported by our subscribers. If you wish to subscribe to Four Media and directly support our work, please go to 44media.o. You'll get unlimited access to our articles if you subscribe to For4Media and ad free version of this podcast.

You'll also get to listen to the subscribers only section where we talk about everyone's story each week. This podcast is produced by Lisa Midcalf. Another way to support us is leaving a five star rating and review for this podcast. That stuff really helps us out. This has been for four four media.

We will see you again next time.