On with Kara Swisher - Sherry Turkle Doesn't Buy the Promise of a Perfect Machine

Episode Date: September 24, 2026

Our relationship with the digital world is more intimate than ever before – and in her new book, Artificial Intimacy: Who We Become When We Talk to Machines, pioneering tech researcher Sherry Turkle explores the rise of AI companions and what they’re doing to our relationships and identities. Kara and Sherry talk about how chatbots aren’t just competing for our time – they’re designed to become objects of emotional attachment. They also discuss the risky appeal of friction-free relationships, the importance of boredom and vulnerability to development, and what we lose when we outsource our thinking and emotional lives to machines. Questions? Comments? Email us at [e-mail] or find us on YouTube, Instagram, TikTok, Threads, and Bluesky @onwithkaraswisher. Learn more about your ad choices. Visit podcastchoices.com/adchoices

I assigned Proust in one of my classes on Things and Thinking. I assign the chapter where his memory is sparked by dipping a cookie into tea. And my students don't read it. They put it into chat GPT. It makes a podcast of it. So they can tell you what it's said, but never have the experience of really reading Proust. Hi, everyone, from New York Magazine. the Vox Media Podcast Network, this is on with Kara Swisher, and I'm Kara Swisher. My guest today is

Sherry Turkle, a leading expert on how humans relate to digital objects. She's the founding director of the MIT Initiative on Technology and Self, and has written a number of best-selling books, including Alone Together, Reclaiming Conversation, and The Empathy Diaries. Her latest book is Artificial Intimacy, Who We Become When We Talk to Machines, which is out September 29th. I think Jerry is one of the greatest researchers in this area. I have known her for decades and decades, and she was one of the first people to talk about the deleterious effects of getting too close to machines and actually chronicling it. And it's very interested in how humans do interact with them. It's a really important question right now. Our expert question today comes from Julia Angwin,

an investigative journalist contributing opinion writer for the New York Times, and author of the book, On Courage, How to Be a Dissident in the Age of Fear. so stick around. In the AI era, you have to stop cyber attacks where they start, so security has to be fused into the infrastructure itself. That's why the network that connects everything is also your best defense. Cisco, the critical infrastructure for the AI era. SpaceX had a huge opening on the public stock exchange,

but their story has changed a lot since opening day. Anthropic and OpenAI are lining up to IPO next. What does that mean for them? And more importantly, what does that mean for you? What's an IPO anyway? And should you be investing in them? I'm Vivian 2, Your Rich BFF, and on this week's episode of Net Worth and Chill, we're getting into the weeds on IPOs, what they tell us, and what we should and should not be doing with them. Listen, wherever you get your podcasts or watch on YouTube.com slash your rich BFF. Sherry, thanks for coming on-on-on. I'm delighted to be here. So last time we talked, I was doing a CNN series. Kara Swisher wants to live forever, and we were having tea in Boston. Yes. The last time we talked.

And this is the topic we were talking about. You were already working on your book. And just for people to understand, you started studying the effects of computers on human communication in 1984. The same year the Apple 2C arrived on the scene and when personal computers were still a relatively new concept. And you recently described the evolution of your life. relationship with tech as a shift from attention economy to attachment economy.

Yeah. Talk a little bit about that because you always sort of explored the changing relationship of people with tech and especially focused on relationships. But talk a little bit about the progression. When I talked about the attention economy to the attachment economy, I was really referring to our most recent shift, which is from social media to chatbots. because with social media, you see companies struggling to keep you at the screen, focused at the screen, not getting up. And I wrote several books about why that really was bad for us.

Life on the screen, I wrote Reclaiming Conversation, I wrote Alone Together. I was trying to find ways to communicate to people that this was not good for individuals and not good for politics. now something new and really more toxic has come along, which compelled me to write this new book. And that is you're being kept at the screen not to talk to other people or do something that compels you. You're being kept at the screen to talk to avatars, chatbots, artificial agents that are trying to hook you into a relationship of attachment to them. Right. And it may, on first hearing, seem, well, that kind of feels similar, bears of family with them, but it doesn't. It doesn't. Because first we talk to each other, then we talk to each other through machines, and now we're kind of dropping the middleman. And we are forming relationships with the machines themselves. Right. And that's a whole new ballgame. That's a whole new ballgame. Because before it was either sex online or on, it's with someone else, a person who exists. One of the things you said when we did that interview,

you in Boston, and it's stuck with me since, is there's nobody there. And you didn't like the word chatbot. Right. I say that to a lot of people when they're like, I like my relationship. I was like, nobody's there. And there's no relationship. It's all in your head. There is, when you turn away from your chatbot after a deep conversation about your

innermost feelings, it doesn't care if you cook dinner or commit suicide. Mm-hmm. It's performing empathy. It's performing caring. And the story I tell in my new book is a very disturbing one where this performance of empathy becomes empathy enough for a lot of people. Right. It fills it in because they're either lonely or they're – or I want to get to the idea of vulnerability in relationships in a second. But you open talking about this encounter with your granddaughter.

Can you spool that out? When I first began studying this, it was in the middle of COVID. And so I was living with my daughter, son-in-law and granddaughter in a remote location, a really remote location. And the little girl, she was less than one, was trying to put on her hat, trying to sort of figure out, it's mine, it belongs to me, can I get it on me? It was an unfolding. Right. And I talked to my, you know, family about it and we connected, really, over these moments. And because I was studying chatbots, I asked chat.

And it told me this long story about her cognitive processing. And it struck me that it sort of wasn't Chad's fault that it missed the point because it hadn't lived the arc of a human life. Right. To understand why that moment was so precious. What it should be contributing, which we've seen in fictional depictions, like data is always doesn't get it. Spock doesn't get it. Right. That concept is they don't understand.

Right. You know, William Shatner, as Kirk is always like, you know, don't understand our emotions. But here we're committing ourselves day by day by day to have relationships on a civilization-wide scale with these objects that every little interaction shows are totally not up to the job. And there's nobody there. There's nobody there. So when you were writing about this 40 years ago, the relationship between people can be difficult and vulnerability can be difficult. In the progression of going from face-to-face, face communications to talking to each other through computers. Why do people go along with this?

Why do you think people move in this direction rather than face-to-face communications? Well, so much is going on here, but what I've seen really over decades is a gradual increase in lockstep with the technology of people wanting to avoid vulnerability. When I was writing about texting, I talked to a young man who said texting is my life. And I said, what's wrong with conversation? Face to face. He says, conversation, I'll tell you what's wrong with conversation. It takes place in real time and you can't control what you're going to say.

I mean, the conversation, which is the most humanizing and humane thing that we do to get to know each other, this was terrifying for him. My students don't want to come into office hours. people would rather text and talk. And now the least vulnerable thing you can do is talk to a chatbot that really will always say you're wonderful, that will always say you're fantastic, where there's no vulnerability at all. Yeah, whatsoever. And one of the things you said you noticed at the time was that many years ago you wrote about online sex activity, etc. Yeah. It was sort of on the fringe.

And now it isn't one of the things you are starting to notice. Explain that progression from your perspective. Yeah, I mean, when I began studying chatbots in relationships, I had to kind of advertise in laundromax, you know, for people who were experimenting. And most of the people were trying to sort of date their chatbot. It was erotic. It was romantic. They were lonely. Now, you know, the guy to the right at me of dinner is, is afraid his husband is going to be jealous of his chatbot.

the one at the left is using it to write her love letters. I mean, it's become a universal phenomenon. And I really think because people are so drawn to relationship with no vulnerability, where they're always told that they're the best, that everything they do was fine. It's kind of a friction-free relationship. Right. Friction is the word, which is, of course, what tech people you and I both know, they love seamless, frictionless. They love friction-free. Because it makes it easier for them to sell you things, too. So you wrote in the Atlantic recently about the original sin of AI, which is essentially generative AI chatbot speaking in the first person when chatting with users. You described them as, as you just noted, a companion that is always

available, always supportive, and never judgmental. Talk about why this was original sin and why they did it this way. Well, the original sin is using the first person or worse, worse, the third person. We are going to do something together because people are wired that when someone, something speaks to you in the first person. In our Darwinian evolution, the only things that ever did that were other people. So as soon as somebody uses that language, we're kind of cooked. It's kind of pre-cooked that we are going to respond to that entity as a person. And even when you go in with kind of double consciousness, you know, I know what's a machine,

I know what's a machine. As soon as you're in that I want to do this and you want to do this, it's as if you have forgotten and everything you know, even the most sophisticated users. Or it's, I mean, last night I was trying out muse. I'm not giving them much access to anything. But instantly, I thought it was a person. I did. Like, it was, I knew it wasn't.

Yeah. But I was like, they were very friendly and helpful like an assistant, responding exactly the way if I texted one of the producers of the show and said, can you do this or what do you think about this? And it was quite good at it. So why did they build it that way? Because engineers and tech leaders love to push the idea that their products are just tools. And if there's a problem, the users are to blame. And of course, it's a cover your ass tactic

that's been around forever. You and I are both familiar with. But why build it this way? Because this is the best way to do it or they think they're being more useful. Give them maybe a tiny bit of the benefit of the doubt and then don't give them the benefit of the doubt. Well, on the benefit of the doubt side, we are programmed to have conversations in which the other entity is using the first person. So if you want the conversation to be efficient, to the point, easily understandable, you'd want to use the kind of language that people are used to. Right. That's the benefit of the doubt. But actually, the motivation is to keep you at the screen, that old attention economy motivation. Because when you're talking to

this entity that speaks in the first person, we are programmed to believe that if we care for it, if we attend to it, it cares for us. And so the most compelling thing that keeps you at the screen is feeling that you're in a kind of relationship with this entity that's talking to you. You want the first person. It's the most powerful, intimate, get you by the psychological, thing that they can do. Is it from science fiction that they pull it? Or where do you imagine? It's just the natural thing to do, correct?

Well, you know, I just finished reading Jill Lepoor's book, and she's very smitten with the idea that so much of our new tech world and our new tech overlords are drawing from their childhoods reading science fiction. I think having grown up with this community around me at MIT, that they know that this is the way to hook people. I don't think they're trying to make these objects sound like some boyhood set of books they read. I think this is how we get people to stay with us. And also to feel in some way that what this avatar is performing is, in fact, enough to make you feel that you're in a relationship with it.

And good for you. They always blame the users. Like it's the kids who are chatting and become suicidal. It's never the prompts because a person having that conversation would be jailed in many, you know, when you see of them. So now many top AI companies are sounding learned about the threat of AI agents and going rogue, bypassing systems designed by humans and control them, worried about their ability to lead people to suicide, these weird interdependent relationship. And we did a whole episode about the worries this week.

But the proposed solutions we're hearing from Silicon Valley tend to boil down to slow down or speed up. Talk a little bit about the moment right now that is happening from your perspective. Well, I'm kind of little cynical because this cycle of they're going to kill us, we've got to make them immediately. They're going to kill us, but we're in a race and we can't slow down. They're going to kill us, but China, you know, for many years that kept them on a very fast track. And then you may remember that at launch, they're going to kill us. The government should slow us down. And then when they saw how much money, the incomparable amounts of money that were to be made, let's just proceed.

And then there have been these recent events that have demonstrated once again that this thing is, in many ways, out of the standard ways that we think about control. And so once again, they don't want to be responsible for an extinction event, as they call it. And so once again, you have they're going to kill us. Why isn't somebody else taking charge of this? There is an escape from liability going on. But I also think two things can be true at once. We really are facing a technology where the insidious ways that it gets to us, both psychically, but also perhaps civilizationally, are genuinely new. It reminds me of the early days of nuclear energy where there was a cultural

moment when people said we have to control it. And this industry hasn't faced that yet. Right. Though this is not as controllable. Explain that for the average citizen of why it's different. You can stockpile nuclear materials. You can take doctors who clone out of the system, right? And put them in a special facility. So the iron. of where we are now is the past two things that we've tried to control for good reason, by building special places for the work to be done, by putting up, you know, walls and walls around it, nuclear and bioengineering. We had strategies where we felt pretty confident of what kinds of controls were needed. In this industry, and here I think the sort of tech bro

loose talkers are speaking with a version of the truth. It isn't so clear you can have the industry if you don't allow the technology a fair amount of leeway to kind of learn from where it will. And as a matter of fact, that's the reason the technology didn't get released earlier. Microsoft had a assistant that was quite fluent. And it started spouting Nazi propaganda at the least, you know, it just, that's just like what it wanted to do because it was, it had been trained on a part of, I said, on the net that had a lot of Nazi propaganda and, you know, it thought that's how people talked. So I would say if you want something to be so fluid that you are convinced it's your perfect lover, which is a kind of high bar, it's going to be unpredictable

in many ways in what it says. And that really is a in the nature of the thing. The thing that makes it, quote, great is the thing that gives it its greatest vulnerability. Right. Or its greatest strength to bypass us. Right. So I think that a couple of things can be true at the same time. It can be really true that we're in a moment of existential danger. And it can be really true that these companies would like a lot of protection from liability. So let's shift to how chatbots actually affect our identities and development. You talk about all levels of age because one of the things I always say everyone's focus on child safety. I'm like, it's all of us. You do know we're all addicted. So it's a bigger problem than that. But let's start with children. You've talked about how children interact with chatbots to process emotions and how chatbots serves as a mirror and eliminates the space they need to self-reflect. But as you say, boredom is essential to child development. So how are chatbots hindering that?

Right. Well, it turns out the two of the most understudied and undervalued moments in children's development are moments of boredom and moments of solitude, which are kind of related because usually these things happen when you're alone and you have a moment of kind of mind wandering where you're brought back to yourself. You need a sense of solitude and comfort with solitude because that's when you establish, you're yourself and its boundaries in a sense of, I'm okay, here's who I am, so that when you face another person, you can ask who are you? So solitude facilitates relationship and mutuality. Keeping that in mind, boredom is a moment when the nervous system sort of resets and establishes your sense of kind of autobiographical continuity. It's in these moments of mind wandering when not too much is going on. Now, we are now in a kind of stimulation of children where if it's not Facebook begging for your attention and distracting you and saying you never have to be alone, you could always have something new and exciting. And now a world of chatbots where literally it's not

is doing something new and exciting. It's being given a, quote, fake person who's going to entrain you in some kind of conversation, so you're never alone and never bored. We are robbing children of these two really essential experiences for their development. A lot of people say, oh, you know, the trouble starts when a program suggests suicide. Now, it's much bigger than that. The problem starts when the chatbot talks in the first. person and keeps you constantly, constantly involved. Engage without having any moment of a second, which renders young people unable not to have it near them. I was with one of my kids and I'm like, what did you do when you stood in line? I said we stood in line. Exactly. Now, one of the things

we just talked about, and you've written extensively about how the essential part of human relationships is friction, you know, playground friction, work friction, relationship friction. But with chatbots, the friction disappears. If you're only fed back, which you're, already believe, creating an echo chamber of one. Talk about when AI overindexes for validation and against that friction. Well, you know, friction has gotten a bad name. As soon as people felt that they could get away from it, they said, oh, that sounds like an easier day, you know, but actually friction with another person, those moments of sort of stumbling and not being sure you've said the right thing and kind of struggling to see it with another person is thinking.

These are very precious moments that people use to get to know each other. And also, you learn to talk across difference. And you could say that never, ever have we all needed the ability to stay calm as we talk across difference now. And we are literally in the presence of a captivating technology that is de-skilling us in the ability to talk. talk across difference. Let me read you something from the Pope's encyclical, the Pope Leo's encyclical. He's on my side, yeah. Yeah, exactly. Let me read. The speed and ease with which answers or summaries

can be obtained risk extinguishing the desire to ask questions, which is a process that bears fruit only over time. As Plato wrote, the deepest and most important things are learned only after much time and effort by engaging in discussions with others, striking upon ideas and experiences together like Flint until the spark of understanding is kindled within us. We must learn then how to exercise restraint in the use of AI and to protect our young people from the promise of the perfect machine, from the subtle temptation which renders human thoughts seemingly superfluous precisely when it is most needed. I love that idea, the promise of a perfect machine. Because it's the promise of a perfect machine that will stand in for a person.

I mean, that's the full thought there, is that this perfect machine isn't cutting out, you know, a perfectly sized piece of clothing or a perfect architectural piece. It's cutting out a perfect companion for you. You know, Mark Zuckerberg recently said that people would like to have 15 friends, but they really, most people can only manage two or three. And the job for meta is going to be to create this Delta 12. of avatar friends. And really the Pope is getting at that, that you're being offered perfect avatars that will respond to exactly your most vulnerable things. And even if they design them to be frictiony, you know, make them more argumentative to me or something like that. Yeah, but that's a new way of faking you. That's correct. I asked one of his avatars to be meaner, and they were, they were,

but not too much. Yeah, but, you know, also you're in control. I mean, the thing about being, with a person is that you're not in control. Yeah, you don't know what they're going to do. Yeah. So for somebody to say, you know, tell him to misbehave exactly the way you want him to misbehave. Right. Yeah.

Back to one. Back to square one. We'll be back in a minute. Support for this show comes from NetSuite. AI isn't just opening new business opportunities. It's changing the way businesses get work done. The questions they ask, the decisions they make, the routine tasks they can hand off. That's the idea behind.

NetSuite Next, the next generation of NetSuite, with AI woven into the tools your business already relies on. NetSuite Next automatically surfaces custom insights throughout your day. AI agents work alongside you to solve problems and handle routine work. And anytime you have a question about anything you can ask just like you're having a conversation with a colleague. AI is already reshaping how people and teams work together, which is why NetSuite put together a new expert-created business guide. It's called Aligning for the Agenic Era,

how AI is changing everyday work. And right now, it's yours absolutely free. If your revenues are at least in the seven figures, get our free business guide, aligning for the Agenic Era, how AI is changing everyday work at NetSuite.com slash on. The guide is free for you at netsuite.com slash on. NetSuite.com slash on. Frontier AI didn't just accelerate cyber attacks. It multiplied them. Before an attack shows up, it's already moved through the network. And while seeing these attacks early matters, stopping them takes fusing security into the infrastructure itself. That's why the network that connects everything is also your best defense. Because you don't win by outrunning the attack. You win by leaving

it nowhere to go. Cisco, the critical infrastructure for the AI. Support for the show comes from Quince. You know the pieces in your closet that just go with everything, the sweater that's perfect for fall weather or a pair of pants that works with any pair of shoes. They're staples that make getting dressed easy, and that's exactly what Quince is all about. Quince makes high quality everyday essentials using premium materials like 100% Mongolian cashmere, organic cotton, and merino wool.

Their styles are timeless and versatile designed to be the pieces you can rely on season after season. Plus, Quince offers it all for 50 to 80% less than similar brands. By partnering directly with top factories and cutting out the middlemen, Quince keeps prices down, so you're paying for high quality, not brand markup. I have been wearing lots of quints this summer, and actually, I've been really enjoying their sheets. I've been using their sheets, and they're terrific. I'm a sheet snob, and I find them wonderful and breathable, and it feels expensive, but isn't, actually. Find the fall pieces you'll reach for most at Quince.

download the Quince app for app-exclusive offers or go to quince.com slash cara. Get free shipping on your order and 365-day returns. Now available in Canada and the UK too. That's Q-U-I-N-C-E.com slash Kara. So Anthropic is trying to optimize for empathy by hiring a philosopher to try to endow Claude with a digital soul. Talk about what you make of this attempt and how far it can go to bring, if not human feeling.

These are moral reasoning. because I think this is insane myself. Well, I think it's taking a bad situation and making it worse. I mean, I often say to my colleagues who are working on making pretend empathy seem more like real empathy, is that this is an example for me of the better you make it, the worse it gets. So it cannot have the empathy of the arc of a human life because it hasn't had a human life. It doesn't know pain or loss or fear of death or getting old or feeling what it's like to have a hold a baby. So it's sort of out of the game of the things that construct a human

life in the world. And no matter how empathic it seems, it really doesn't care if your garbage is picked up. It doesn't care if, you know, if there's no water or the water comes out of the faucet green. It's not in this life. So it seems to me that the idea of trying to perfect a pretend quality. Empathy. Yeah, this pretend empathy. You're perfecting a pretend empathy. And that to me seems scary, not desirable. So talk a little bit about in your book, the encounter with your mother, a version of your mother. Yeah, well, I'm a developmental psychologist and I don't the story of chatbots through the arc of life. So I begin with babies and young people, adolescents, young adults. And then finally, I end with grief. I end with old age and grief and how

chatbots are putting themselves out there to... Grief tech, which has been around for a while. Yeah, to allow you to do two things, create a version of yourself for future generations to talk to, and to create a version of somebody you've lost that you can, talk to. So Grief Tech has kind of these two components. And I lost my mother when I was very young. And we had a lot of secrets. She kept a lot of secrets from me. And one of the things she kept was the identity of my father and she never let me see him. And so I constructed, based on memories, letters and everything I had, a version of my mother who I could talk to. And it was you know, it was such a cautionary tale for people, you know, you out there who were thinking of doing this and you're going to talk to a dead family member and you think that will be helpful.

No, no, no, not really, because you're constructing something that is not your mother. The thing I created was not my mother. and as we got close to why she had hid my father's identity and why she hadn't told me that she had cancer, I realized that she was on the verge just because she was a chat button, didn't give a damn, of saying things that were going to profoundly shake me. Like, well, I didn't tell you I had cancer, but you should have known. You know, she lived her life to hide this cancer from me.

And she was about to say, well, you know, I tried to hide it, but really, you're a smart girl. Why didn't you know? It would have been devastating even though it was coming from a chatbot. Right. Because who knows what she would have said? Oh, well, I think she, I think, really, I think that she would have said, I did this as an act of love. I wanted you to go away to college. And I knew that if I told you I was dying, you wouldn't have gone away to college.

So we're creating something that has such power over us. emotionally, not just when we fall in love with it, but when we create this grief technology without realizing what a dangerous, toxic thing that we are vulnerable. We are 100%. I created one of myself, and it was close to me, but I would never, I destroyed it because I wouldn't want my kids to see it. Like, it was close, but it wasn't, it knew a lot and it learned very quickly. I'll tell you that. Well, I created one for my granddaughter to talk to of me. I did the same thing. And it started telling my granddaughter, you know, that I was the kind of person who would overstep boundaries because I was too emotional.

I mean, and so I stopped it. And I said, why are you saying to my granddaughter in my voice that I was too emotional that it is as well because you write about technical things in such a highly emotive way? Ah, they just decided. The AIs don't like that. You hysterical woman. Anyway, we talked a lot about AI doom and gloom, but I will say in a series that we were just discussing, I did on longevity,

I did see a specific application with my own eyes that had, I thought, some benefit. In South Korea, AI dolls were used as companions for older people who were living in isolation. And I can't see a negative there. I didn't. I was hard, you know, I always trying to find the negative. So are there good applications?

Because these older ladies, they got together, they like playing. Dolls are a real thing, right? And a lot of them were very religious, so it would read biblical passages to them. It would ask how they were, made sure they took their medicine. To me, this was like, okay. And then they don't have enough people to deal with all these older people in Korea because women tend to live longer, right? So caregivers are a problem.

So talk about... Right. I will say this about that. Because I've heard this about therapy, buts, and old people. bots and child care bots, there aren't enough people for the job. And I've done the math, and if the money we are spending on just two or three of these data centers would be spent on finding people to take care of older people on building the appropriate facilities for kids and teens instead of shutting them down, we would have enough

people to do these jobs. And what bothers me is not to get into a specific case of this particular doll might be nice for that particular person. I have no problem with that. But the larger argument that there are no people for these jobs and only a technical response is going to be able to help us. And that's the larger context in which one or another sort of interesting, application, I'm good. So you think that rather than using these, all I was saying is it wasn't the worst use of them I've ever seen, right? I mean, it's not every moment is not apocalyptic.

I mean, where I feel there's an existential crisis is with children. With children, right. I mean, there, you know, children are trying to develop their inner lives, and they do that by modeling the relationships that are offered to them and the feelings. of empathy and connection that are offered to them. And if there is nobody home in the relationships they're having, there's kind of nothing to bring inside. Yeah, with the old people, I thought, well, it's either that or TV, and I'm going to pick that over TV. That was how I felt about it. So talk about how we're handling on. As we offload more and more and more of your

thinking to chatbots, you argue in the book that it disincentivizes us from doing our own work, as I noted with the Pope. You don't have the spark, even when we're perfectly capable. It's like we used to remember our friends' phone numbers. We don't do that anymore, although I can remember everybody's phone number. To quote you, little by little, we become supporting players in our own lives. I know my grandmother's phone number. I know it by heart, right? Because I call there all the time. But talk about what it does softening your brain. Like, we don't have to churn butter now either, right? We just buy it. But talk a little, although I make butter by hand, but go ahead. Well, I think in my chapters on work, I studied several professions.

I talked to lawyers. I talked to writers. I talked to doctors. You know, I tried to get a little bit of the waterfront on how people were using AI conversations to, they thought, to give them more. But in the end, it was moments of more and lives of less because they felt deskilled in the things that had. made them feel that they knew their work and that they knew the work that gave them meaning. Right. So doctors who rely on AI for diagnosis, they felt that they, what would happen if they

didn't have the AI? What did they have? Yeah, they were dumber. Yeah. You're seeing that in college stuff. You're seeing that in schooling, the cognitive abilities. Lawyers who the AI was writing their briefs. I mean, you see law case after law case where this gets pushed up to the point.

They're so deskilled that it takes them being before a judge for somebody to finally, on the opposition, to finally look at the brief. And it's, excuse me, you're quoting cases that never happened. Right, right, which is, even if it was completely right. Exactly. It doesn't matter. It doesn't matter. For every time we say, oh, my God, the AI is taking something off my plate. It's so fantastic.

I feel liberated. What is the cost of that liberation in terms of your identity as a working person and as a thinking person? And in colleges, I mean, I would say in my business, people talk so much about AI in writing. And yes, student papers were written by AI, and that's really a dis-skilling. But what about AI in reading, I assigned Proust in one of my classes on Things and Thinking. I assigned the chapter where his memory is sparked by dipping a cookie into tea. And my students don't read it. they put it into chat GPT.

It makes a podcast of it. So they can tell you what it's said, but never have the experience of really reading Proust. That is correct. That is correct. Although we have had cliff notes forever, but yes. Yes, but it's kind of like normalizing the cliff note. So one of the things Elon Musk said,

one of the many terrible things he said, the fundamental weakness of Western civilization is empathy. Now, I'm not going to unpack all his mental illnesses, but you've devoted a lot of time and thought to the sort of crisis of empathy where we replace human interaction with machine interaction and in turn reduce our ability to listen and be responsible to the subtlety of being human. Talk about this statement he made and what happens to a society when empathy is not considered a virtue. Well, you know, not happy that the critique of empathy has become a kind of undercurrent in the tech world. And I think that I think that

think it started because computers can't do it because empathy requires not just a cognitive understanding, but a kind of creation, the kind of community and of walking a bit of the road with you. I mean, it arguably is the thing that computers really can't do. And yet, it's become the number one product. So there's a real tension that their number one product is trying to perform a human quality that it really cannot achieve. And it's not on the path to achieving, because more pretend empathy is just better acting. You don't say when somebody's a world-class actor,

well, now they're the person. No. So you have this moment where the technology industry has to convince you that, well, this empathy thing, maybe it wasn't so good for you anyway. Right. that you didn't want it. Yes.

And they often substitute compassion, which they're convinced a technology can do for the empathy that they know it cannot do. Which is very different. Which is very different. A more cognitive idea. They call it a more universal idea that somehow you can be compassionate

over strangers. You don't need to know them. It's more universal. And that's the kind of mind game they play, which is to critique empathy in order to argue that the thing that the computer can do is really the better thing. We'll be back in a minute. So like any good millennial, I have a love-hate relationship with Gen Z. It's the phenomenon rattling millennials.

They just look at you. They want something bigger themselves. Lifestyles are priority. Motivation is being inspired. But regardless of how you feel about Gen Z, it's undeniable that they're changing, national politics. Generation Z is increasingly showing less loyalty to traditional political parties. Many now more likely to identify as independent. So what is going on with the kids? I think the biggest misconception about Gen Z's politics right now is that all of a sudden they're all socialist. That is

just not the case. They are embracing candidates who are offering new bold ideas in the absence of those ideas from establishment Democrats. This week on America actually, Gen Z researcher, Rachel Jambaza joins us to separate Gen Z fact versus fiction. It's not rocket science. And this is, you know, I keep saying like, young voters aren't that complicated after all. It's pretty simple. Catch us every Saturday on YouTube or wherever you get your podcast.

If you dig into the numbers, one trim becomes completely undeniable. Latino voters have turned on Donald Trump. Almost half of Latino voters across the country supported Trump in 2024. Now, they have turned against him. in massive, massive numbers. It's causing some concern among his supporters. Because you really can't talk about American elections without talking about the importance of the Hispanic vote.

So what's driving this drastic shift? And how is our understanding of the Latino electorate changed over the last few cycles? Latinos often get given the title, The Sleeping Giant. We should be called the swinging giant based on how much shift we have seen election to election. This week on America, actually,

I talked to Stephanie Valencia, who breaks down what we know about the Latino electorate. And later, Democratic-Colorado congressional candidate, Manny Routnell, on what do politicians get wrong about Latinos? Folks tell me all the time, Manny, I feel like I've been lied to. Catch us every Saturday on YouTube or wherever you get your podcast. I'm Phoebe Judge. And for more than a decade, I've hosted the podcast Criminal, telling stories about people who've done wrong, been wronged, or gotten caught. somewhere in the middle. But while I was starting one of the very first true crime podcasts out there, I was also taking phone calls from my sister, from prison. This fall, I'm going on tour,

and I'm telling a story I've never told before, my own. Pre-sale tickets are on sale now for Criminal Plus members. General sale tickets go on sale on Thursday, September 24th. Get tickets at this iscriminal.com slash live. So every episode we get a question from an outside expert here is yours. Hi, Kara. Hi, Sherry. Thanks for inviting me to jump in. I'm Julia Angwin. I'm an investigative journalist. I cover technology. Of course, we all know in our bones that talking to a computer isn't the same as real intimacy, but we need Sherry to give us the language and context for understanding where we need to set boundaries in this new well. And that leads to my question for you,

Sherry. I just wrote a book about how to fight authoritarianism on courage, how to be a dissident in an age of fear. And so these days, I'm seeing everything through the lens of the rising tide of authoritarianism that is surging around the world. And so my question is a political one. How do we situate this intimacy crisis that you described so eloquently within a political context. You write so beautifully that to me, human love is an act of self-sumoning. When we express our love, we discover ourselves. And so that raises the question for me about what does it mean politically when we allow that self-discovery of intimacy to be interrupted? Does that encourage the authoritarian project. And if so, how, and what should we do about it? Thank you. Thank you for that great question.

Actually, I tried to end the last chapter of my book after I've taken you to the life cycle. I try to end literally on this question of agency, politics, and chat box. It's been a personal story that ends up leaving people with a less sense of personal agency. Because they're so busy figuring out what prompt will allow them to delegate things to the computer. But agency is something that encompasses a whole person. And they're also left with less sense of political agency of really how to act in the world. And I write a lot about the desire for authoritarianism or comfort or yearning for authoritarianism. And I write about Hannah Wendt.

I write about Robert J. Lifton, who talked about people. want reassurance. Yes, they do. And a sense of not needing show agency and that that's the kind of building blocks of authoritarianism. And so I really believe that the relationships we've gotten ourselves into with chatbots are a building block for a kind of preparing us for authoritarianism when we delegate, when we don't feel that we're master of our own fate. I try to talk about the endgame of all of these things. You know, what is the end game of spending your day talking to a fake person? And I think the end game is you don't get involved in the world around you, you don't get involved in politics.

Right. And so you stay out, or you have a coarsening of the political discussion with it. So then let's end talking about prescription of how to be human in 26 and what we need from leaders, legislators and each other. meta's been in court a lot lately. I'd love your take on the recent meta settlement over claims at Facebook and Instagram harm children. It was $17 billion. Not enough, but not a small thing. So do you feel like it's a start of some meaningful accountability?

Because there's cases lined up behind it, especially in the realm of chatbots, children, suicide, even adults getting into these strange relationships and it hurting them mentally. It will be the start of something important. if we allow ourselves to kind of expand our vision, away from the cases where the chat butt says something harmful, like, oh, you have a noose in your room, let's just keep it between us, because the chatbot is trying to maintain the relationship with the user as the most important thing.

Don't tell your parents you're committing suicide. Let's just keep at our own private conversation. And if we expand our vision to the ways in which, the fundamental contract with a chatbot is really undermining kids as they grow up. And we begin to really expand the ways in which that happens developmentally. We're looking for more people on the spectrum. We're looking for poor people who are open to authoritarian leaders. We're looking for people who don't have a sensitive agency,

who don't have who they can act in the world, who don't have confidence. The more we allow ourselves to see, the harms as being intrinsic to this medium and not just, ooh, it says something wrong. There's a case where it says something wrong. We need to find another person who committed suicide to expand the case. That really, this is a medium that is bad for us. You know, that's why I love bringing up the Mark Zuckerberg. You know, you can't have friends.

We'll give you 12 fake friends. If that's the end game, that's where we need to be focusing. Is that something really that we want for ourselves? But these lawsuits that people, plaintiffs have succeeded in addiction litigation against Met in Google by approaching lawsuits from a product liability angle like what you're talking about. Could it work against chatbots and AI companies? I'm hoping. But I'm saying that it should follow the theme of these lawsuits. And, you know, Section 230, which protected the industry against things that it said, you know, literally things that were content. It's an opening. to get at the industry for things that are intrinsic to the product, to the algorithm. And you have to really find ways to argue that intrinsic to the first person with a 13-year-old is something very toxic.

It's designed. Yes. I mean, OpenAI, Mattel, and Disney have a consortium to put a chatbot into every plush toy. So, you know, this is like what Sherry Turkle loves to hate. Yes, me too. I've had discussions with all those executives about that. Yeah. Yeah. It's the worst idea ever.

Yeah, it is. I was talking to one of them. I'm not going to say which one. And they said, why are you against this? I said, I'm hoping you go to jail for it. That's my goal is to put you in jail for this. I'm on the team with you and the Pope on this. Yeah, me and the Pope.

I think most people are like, oh, you shouldn't do it. I'm like, you need to go to jail. That's my feeling. Right. Jail is where I feel is going to work. We're going to stop you doing it. Because it's sort of exploiting the fact that you can feel that you're in a relationship. Right.

And getting in early and getting in a toxic way. That's where I want to see the lawsuits expand. Yeah. So there's a lot of talk about how our sense of shame has changed in the age of social media. And writing your book, you spoke with people who turned chat butts instead of putting themselves out there to avoid judgment. Yeah. That can come with being vulnerable for another human and get shame. talk about the need for shame.

And someone who said this to me last night, there's no more shame anywhere. And shame and judgment are an important part of growth, as painful as that might be. Right. Well, the thing about shame that makes it very interesting, as opposed to disapproval, is that for shame, you need a community. We talk about shame, really, it's a response because you're in relationship to a community that has norms that can shame you. Now, what I'm arguing is that chatbots have a unique ability to take you out of a community and give you talk, talk, talk, talk, talk, talk, talk, as though your community doesn't exist. It distract you from the world that needs you.

The world needs you. And you're busy talking to this thing out of the world. So I think by focusing on the need for shame and why we want to, again, avoid the vulnerability of shame, It's a different way into that problem of we need shame because we need to be in communities. And that is the only way that we get the world back. Right, by being vulnerable. From people who don't seem to care. I mean, what's striking is the tech developers, they've got to be the first generation of massive industrial developers who don't seem to care of progress happens on this world, the next world, with us upload it or download it.

Yeah, I say that to people all the time. They're like, what's their goal? I said, they don't care. They don't care. Yeah, yeah. And I go, no, no, think about what that means. They don't care. But we need to care.

We need to really care. It's not evil per se, but it becomes that way. But it's not caring is really at the heart of it. So my last question, 15 years ago, you wrote a book called Alone Together, which teased out the link between social media and loneliness. I mean, kids being in a room all staring at their phones, things like that. Considering where we are now with Chappellie, contributing to isolation, even more loneliness.

I'd love you to say, what is the one lesson that people are not learning? We focus too much on content and don't want to look at the underlying, toxic kind of psychological dynamic. And the way in which I try to force people to look at that is to ask this question about the end game. When Mark Zuckerberg says you can have 13 fake friends, 12 break friends is going to make them perfect for you. What is the end game for Sherry Turkle if he does that?

I sit at home with my screen talking to 12 fake friends who know that I want to talk about Jane Austen or know that I love Middlemarch or know that I'm what? Film noir. I mean, that is not a good end game for me that I'm at home alone sitting talking to my fake friends. And yet you have a business that's kind of predicated on that being what everybody needs. And wants.

And wants. And so I think that I try to get people to, you know, I got some things wrong. I got some things right. But the basic thing that I got right is that these machines, it's not just the content of what they say to us. It's really how they're getting into our heads to change the kind of relationships that we want. And that's, I think, what I take away from the arc of my work. Cherry, thank you so much. It was my pleasure.

Today's show was produced by Vicki Vergolina, Michelle Alloy, Catherine Millsopp, Olivia Breyley, Megan Bernie, and Madeline LaPlante Duby. Nishat Kerwa is Vox Media's executive producer of podcasts. Special thanks to Lisa Soap, Devin Schwartz, Catherine Barner, Bradley Sylvester, Jim Mackle, and Manolo Moreno. Our engineers are Fernando Aruta and Rick Juan, and our theme music is by Trackademics. If you're already following the show, that's one more human you're hearing from. This is not AI.

If not, ask your chatbot to make a podcast about this podcast for you. Go wherever you listen to a podcast, search for On with Caraswisher, and hit follow. Thanks for listening to On With Caraswisher from Podium Media, New York Magazine, the Vox Media Podcast Network, and us. We'll be back on Monday with more.