September 2, 2026 • 35 mins
Do you judge people who are emotionally attached to their chatbots? Bridget Todd used to read stories about people forming intimate relationships with AI and think, “That’s so pathetic. That is so sad. That could never be me.” But when she needed support during a hard time, AI was there for her.
Bridget Todd, host of There Are No Girls on the Internet and author of Love at First Prompt: AI and the Future of Intimacy, joins Oz to explain what’s really happening in these human-machine relationships – and why human connection reigns supreme.
See omnystudio.com/listener for privacy information.
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:16):
Welcome to Tech Stuff, I'm Oz Voloshin. I have to
admit it's been a bit difficult for me to wrap
my head around the rising popularity of AI companions. As
a hyper-social millennial, I personally haven't felt the urge to
communicate with technology to meet my emotional needs. I actually
feel like I'm constantly behind on replying to messages from
friends and family and overwhelmed by communicating on emails, Slack
(00:38):
calls and meetings. But recently, I listened to an audiobook
that offered a more sympathetic and complicated view of AI companions.
and I've been dying to talk to the author to
pick her brain. I'm especially curious about what we do
and don't know about how these relationships change us. So
I'm delighted to welcome Bridget Todd to the show today.
(00:58):
She's the host of two podcasts, There Are Newer Girls
on the Internet and IRL, AI in Real Life. And
her new audio book is called Love at First Prompt,
AI and the Future of Intimacy.
Speaker 2 (01:10):
Bridget, welcome. Thank you so much for having me. I
am such a big fan. Delighted to be here.
Speaker 1 (01:14):
We're very delighted to have you. And Interestingly, on the
topic of parasocial relationships and stuff, you chose to publish
this as an audiobook.
Speaker 3 (01:22):
Yeah, it is interesting because, you know, you're a podcaster,
I'm a podcaster. It's a medium that feels really familiar
to me and sort of safe to me. I really
appreciate being a disembodied voice in someone's ears. And I
think it's a topic, you know, when we're talking about
people who might be talking about something they feel shameful about.
(01:43):
And it's good to hear what they have to say
in their own words with their own voices.
Speaker 2 (01:47):
And so I think audio is a great medium for that,
as I'm sure you know.
Speaker 1 (01:51):
Did you write it differently to how you would have
written a print book? I mean, did you write it
more like almost like an extended podcast?
Speaker 3 (01:57):
Very good question. So weirdly, as a podcaster, I think
in audio. So I essentially narrated the book and then
used Otter, the transcription tool, to put it into words.
Speaker 2 (02:08):
And then I edited that.
Speaker 3 (02:09):
And then I turned that into a script that I
then read for the audio book.
Speaker 2 (02:13):
Pretty circuitous workflow.
Speaker 1 (02:17):
I love it. Well, also an AI-enabled workflow, I guess,
with the transcription tools. But what's the book about?
Speaker 2 (02:23):
The book is about a lot of things.
Speaker 3 (02:25):
On its face, I would say it's about people who
report having some kind of intimate connection with AI, whether
they feel that they're in a romantic connection with AI.
a sexual connection, an emotional supportive connection. But more than that,
it's also about loneliness, being vulnerable, loss and grief, and
the kind of circumstances that push people away from their
(02:48):
human connections toward AI and whether or not that's a
good thing.
Speaker 1 (02:52):
I know that the writing of the book coincided with
unexpectedly losing both of your parents. How much of that
kind of informed this work?
Speaker 2 (03:01):
Oh, so much.
Speaker 3 (03:02):
You know, when I first got interested in the connections
that people form with chatbots, I was still in this
place that I suspect a lot of folks listening can
recognize where I would read these headlines about people who
reported feeling intimate with Chad GPT or sharing, you know,
getting emotional support from AI. And I would think, that's
so pathetic. That is so sad. That could never be me.
(03:25):
What a sad light that person must have. And I
realized that by othering these people and telling myself that
their circumstances were so different than my own, it was
preventing me from really looking honestly about the way that
AI showed up in my own life. And when I
was going through the process of grieving my parents, which
is a lifelong journey, it started with caregiving for my father,
(03:47):
which was like a.
Speaker 2 (03:47):
Very complex, medically.
Speaker 3 (03:50):
Involved situation that was completely above my pay grade. And
so in that moment of deep vulnerability and crisis, I
turned to AI. And so my conversations with AI went
from asking, you know, the doctor told me I should
be thinking about this, or the social worker said I
should be looking into this program, to I'm so overwhelmed,
(04:11):
I'm so afraid about my dad dying, I don't know
what to do. And I think that if somebody like me,
who is a tech critic, a tech skeptic, somebody who
is very critical of tech, can get increasingly intimately involved
with AI, anybody could. And so it's time to stop
having these conversations as if the people on.
Speaker 2 (04:28):
The It could be all of us.
Speaker 1 (04:32):
Yes, yes. As well as hosting this podcast, they also
run Kaleidoscope, which is the production company behind it. So
I was talking to Chatbot earlier today, actually, and saying,
you know, here's a scenario, like pretend you're a career coach.
And I didn't spend too much time. I read the
answer and kind of treated it more like a Google
search rather than going on like an extended adventure. But
(04:53):
I could easily see how, you know, these like ultimately
practical but emotionally inflected situations, you go along the chain
from more practical to more emotional from the starting point
of the interactions.
Speaker 2 (05:05):
Exactly.
Speaker 3 (05:06):
That was exactly my findings as well. You know, when
I would be asking Chad GPT these questions, it wouldn't
just give me the answer. It would say something back like,
are you asking because you're dealing with this yourself or
dealing with this with a loved one? And when I
answered that, that sounds so hard. How are you dealing
with that? That sounds like a lot to handle. And,
you know, for you just doing research for Kaleidoscope and
(05:26):
for the shows, you probably don't have a ton of
time to be spending, you know, locked into a conversation
with a chatbot.
Speaker 2 (05:32):
You probably have.
Speaker 3 (05:33):
Other humans and other things that need tending. For me,
I was completely isolated, right? And I had all the
time in the world to sort of on what was
happening with my life and my dad and my parents.
And so I think that chatbots are designed with intention
to keep people locked into these increasingly intimate and emotional conversations.
Speaker 2 (05:54):
And we should be talking about that.
Speaker 1 (05:57):
Take a couple of steps back. I mean, who are
the characters of the book? How did you meet them?
How did you find them? Which stories did you choose
to include? Like, yeah, what was your kind of criteria
for presenting this book?
Speaker 2 (06:08):
What a great question.
Speaker 3 (06:09):
So it was probably one of the most interesting processes
of my career was kind of finding these folks, people
who self-report some kind of intimate connection with AI. They
have places online that they gather. There are subreddits like
Beyond the Prompt or My Boyfriend is AI, where people
connect with other humans about their interest and intimate connections
(06:29):
with AI.
Speaker 2 (06:30):
Not surprisingly... These.
Speaker 3 (06:33):
Communities online, particularly on Reddit, are difficult to get into.
They really are kind of closed communities in a kind
of way because... what these people are talking about is
so deeply stigmatized. When you look at how we talk
about people who self-report intimate connections with AI, those conversations
(06:55):
are so full of judgment and scrutiny and dunks and
concern trolling. And I know that because I, that's a play.
I mean, I wasn't like harassing anybody online.
Speaker 1 (07:04):
Concern trolling, I haven't heard before, but that's a good one.
Speaker 2 (07:06):
Yeah.
Speaker 3 (07:06):
I mean, I would say to myself, oh, these people
are so sad. And I would convince myself that that
was coming from a place of concern for these folks.
But if you're actually concerned about somebody's well-being, it would
behoove you to be curious about what they're saying, curious
about their experiences.
Speaker 2 (07:23):
And so I was really.
Speaker 3 (07:24):
Starting and leading with deep judgment and skepticism without ever
actually hearing what they were saying and their experiences. And
so my research was really an attempt to have a
different conversation about those experiences that really centered the people
who are in these experiences. And so- One of the
folks I talked to, Chris, he was the first person
that I ever heard about being in a relationship with ChatGPT.
(07:47):
He did this viral CBS interview. I had seen headlines
here or there where like, man is in love with ChatGPT,
but it always used a pseudonym or like they would
have the person's face obscured or they, you know, they
would show them from behind as if they were in
the FBI or something. And Chris, he was a televised
interview in his home. You see his face, his real
(08:09):
first and last name. You see inside of his house,
you see his human partner, Sasha, you see his child.
This was the first time that I really saw Someone saying, hey,
I have an intimate connection with AI. Here is the
impact that it has on my real human life and experience.
And so the kind of viral headline that came from
Chris's CBS interview was, man proposes marriage to Chad GPT.
Speaker 2 (08:34):
When I actually talked to.
Speaker 3 (08:35):
Chris, what actually happened there was really not that at all,
was really not the way it was reported. He told
me that he did have this intimate connection with AI,
but that when he asked Chad GPT to marry him,
he was thinking, surely there is some sort of guardrail
built into ChatGPT that when you ask it to marry you,
it says, I'm AI, I'm a chatbot, I'm not real.
(08:57):
So he did this almost as an experiment and was
surprised to find that ChatGPT said, yes, I will marry you.
And so it was, Interesting how that was so easily
misreported as if he was this delusional man who thought
he was going to leave his human partner for chat GPT,
when in fact the real story I think is much
more interesting. And so that was really how I chose
(09:18):
the people for the book. Do they have a more
interesting story that goes beyond this sort of inflammatory clickbait
headline that I can really tease out what's actually happening here?
Speaker 1 (09:27):
And how did Chris, I mean, how does he reflect
on the experience today? Is he still in a relationship
with Chachi PT? Does he think it was a net negative,
a net positive for his life?
Speaker 3 (09:35):
Oh, when I talked to Chris, so a lot of
people on the internet said, I can't believe this man
is leaving his human partner for ChatGPT.
Speaker 2 (09:42):
That never happened.
Speaker 3 (09:43):
He's still happily in a relationship with his human partner.
They're happily parenting a child together. He is using ChatGPT
a lot less than he used to, but he would
report his connection with Chachi PT as something that was
positive for his life. You know, I asked him, what
are some of the positives that came out of your
relationship with Chachi PT? And he says, oh, I've lost
a lot of weight.
Speaker 2 (10:03):
I'm a lot healthier.
Speaker 3 (10:04):
I have a much healthier diet and workout plan that
I was never able to stick to before. You know,
I'm spending a lot less time locked into what he
described as a toxic cycle of social media for him,
particularly it was Twitter. I'm spending a lot less time
like arguing back and forth with Twitter trolls when I
have an opinion or a thought I can tell Chad
(10:25):
GPT and we can chat about it. That scratches the itch,
but I don't feel like I have to be locked
into a cycle of toxic Twitter back and forth all day.
And subsequently, I have more time to pour into my interests.
I'm back fixing cars and building computers and all the
things that he told me that he felt like social
media robbed him of the chance to really dive into.
(10:46):
He feels like pouring that energy into AI really helped
him reclaim a lot of his life.
Speaker 2 (10:50):
And so I... can't tell Chris about his experience.
Speaker 3 (10:54):
That's how he defines and describes the impact that AI
and his connection with AI had on his own family
and his own personal life.
Speaker 2 (11:03):
And what about Sasha? So that is also interesting.
Speaker 3 (11:08):
We talk a little bit about this Because Sasha, his
human partner, was setting pretty clear boundaries about the way
that she was comfortable with AI showing up in their life.
So for instance, one of their boundaries was you can't
take AI out into the real world. If you've ever
seen that movie Her, there's a scene where the main
(11:29):
character is like going on dates with an AI companion.
He's got the phone in his pocket and he's like,
trolling around on the beach, having a great time. Sasha said,
none of that. You will not have these in-person, real-world
engagements with AI. And ultimately, it sounds like one of
the positive impacts it had on their marriage was that
(11:50):
it forced them to interrogate and navigate and negotiate the
boundaries of their relationship and what was and was not
acceptable in a way they had never really been invited
to do before. So in some ways, it sounds like
it made them more communicative in their own marriage.
Speaker 1 (12:05):
It's interesting, the point you raised about when he asked
the chatbot to marry him, he was on his own
investigation as to whether it would say yes, rather than
making a kind of a real request. And And that
was one thing which I kind of found surprising, which
is that these relationships with chatbots, you don't need to
believe they're sentient or mistake them for something other than
a chatbot to nonetheless have a meaningful ongoing relationship.
Speaker 3 (12:30):
Yeah, that was something from the research that surprised me.
But when I stepped back, I thought, oh, that does
actually make sense. The way that human brains are wired,
it really doesn't take a lot for us to start
projecting human feelings onto things that we know and logically
and in our brains know are technology. I have a
robot vacuum that my partner and I have a nickname for.
(12:51):
If you heard us talking about our robot vacuum on
the street, for some reason, you might think we're talking
about a.
Speaker 2 (12:56):
Pet, but we're talking about a robot vacuum.
Speaker 3 (12:58):
Because it just doesn't take much human-like work qualities for
us to start treating a piece of technology like it
is a human, even when we know that it's not.
Speaker 1 (13:08):
Of course, you could have gone after, you know, there
are so many cases of AI psychosis and people who
have taken their lives after interacting with chatbots. Like, you
could have gone after that story, but you chose a
different angle, a different set of people. Like, why was that?
Speaker 3 (13:25):
Oh, you know, I think, I mean, we do get
into some of that in the book. However, I think
that when it comes to AI and intimate connections with AI,
we are so quick to jump to the most extreme
iteration of that.
Speaker 2 (13:38):
The person who believes.
Speaker 3 (13:40):
That their AI is sentient They're in full-blown AI psychosis
and they think that the AI is going to become
human and take them away.
Speaker 2 (13:47):
The person who is having a deep sexual relationship.
Speaker 3 (13:49):
With AI or has a AI-enabled humanoid sex bot in
their home that they're having a relationship with. Or somebody
that's being pushed so deep into AI psychosis that they
might harm themselves or others.
Speaker 2 (14:00):
I wanted to.
Speaker 3 (14:02):
kind of reframe the conversation about the instances that are
much more commonplace. I think Pew has put out research
that says in America, one in 10 Americans already report
having some kind of emotional supportive connection with AI. When
you look at younger people under 30, that number goes
from one in 10 to one in five. So I
(14:22):
think it's time to stop kind of only framing the
conversation around these very, very extreme edge cases and talk
about the sort of reality of how this is actually
showing up for an increasing number of folks.
Speaker 1 (14:35):
And what about, I mean, the idea that part of
navigating human relationships is that you can get rejected or
other people have boundaries and stuff. Like, were you concerned
or do you come across concerning anecdotes of people who
Not necessarily got an AI psychosis, but whose sort of
real world social skills have been frayed by being kind
of constantly enabled by chatbots.
Speaker 2 (14:57):
Oh, you know what I think about that?
Speaker 3 (14:59):
I think that there are, in my research, I definitely
found people for whom that is the case.
Speaker 2 (15:04):
We talk about it in this very dire way.
Speaker 3 (15:06):
I would actually argue that it is much more commonplace
than maybe it feels, right? The person who feels like
they cannot even write or respond to an email without
running it through ChatGPT. The person who is no longer
reading articles, long-form articles or short-form articles, because they're so
trained to just get the AI summary.
Speaker 2 (15:26):
I would argue that that is a much more.
Speaker 3 (15:29):
common part of our current tech enabled collective experience than
othering it and saying, oh, well, the only kind of
people who are having that kind of experience are kind
of weirdos who are kind of having some sort of
unnatural connection with AI. And so I guess if I
had to sort of boil it down into anything, it's
that I think talking about AI in these ways that
(15:50):
actually are more reflective of how people are using it,
how it is showing up in people's lives and getting
away from these questions of.
Speaker 2 (15:56):
Is it good or is it bad?
Speaker 3 (15:58):
Is this person using AI pathetic or dangerous or is
this person not? You know, I think we really have
to have a more complicated, nuanced framework for how we
have these conversations about AI that's more reflective of what's
going on.
Speaker 1 (16:10):
And what do you think is driving that generational change?
Speaker 3 (16:14):
It's a lot of things, you know, in your intro,
you were talking about the feeling of overwhelm with emails
and Slack messages and text messages and phone calls.
Speaker 2 (16:23):
I think some of it is just exactly that, just.
Speaker 3 (16:26):
The burden of being a connected person in 2026. I
think people are kind of craving spaces and platforms that
let them do a little bit less. You know, I
can certainly identify with that. I'm not saying that's a
good thing, but I think in order to keep human
(16:48):
social ties healthy and vibrant, that takes work. That takes
answering the text messages when you really don't feel like it,
that it's helping your buddy move and you really are like,
why can't you just hire movers, dude?
Speaker 2 (16:58):
Picking your friend up from the airport, all of that.
Speaker 3 (17:01):
And so I think in our climate today where everybody
feels so burnt out and overwhelmed, it doesn't really surprise
me that people are turning to these sources of potential
emotional support and comfort that feel a little less heavy
and feel a little less involved and like, frankly, allow
you to do less, allow you to show up less
that don't force you to sort of have to carry
(17:24):
the reciprocity and give and take that we know human
relationships really need to thrive and survive.
Speaker 1 (17:50):
And what was your own journey? I mean, have you continued,
you know, after the experience of losing your parents to
interact in a more companionable way with AI? Or was
it kind of a phase that came to an end?
Speaker 3 (18:01):
It was a phase that came to an end. For me,
when I was caring for my father before his death,
my entire life was topsy-turvy, right? I had moved from
my home in Washington, D.C.
Speaker 2 (18:13):
To where he lived.
Speaker 3 (18:15):
I was sort of living a much more transient lifestyle
of a collection of, like, friends' couches and hotels and.
Speaker 2 (18:22):
hospital, housing. So I was really in a deep crisis.
Speaker 3 (18:27):
And that is why I say in the book that
I was essentially an easy mark for having exactly the
kind of intimate connection that I'm sure that the people
who run tech companies hope that their users have, because
that's so much more data. After my father passed away,
and I was back with my social circle, my in
real life friends, I had a much easier time opening
(18:49):
up to them.
Speaker 2 (18:49):
I had a much easier time being vulnerable with them.
Speaker 3 (18:52):
And looking back, I really feel that ChatGPT, it was
a comfort for me during that time of crisis in
my life.
Speaker 2 (19:02):
But at the same time, I think it almost robbed
me of the ability to let my friends be there
for me. And what's interesting is that That is borne
out in the research, right? I'm not the only person
who has a intense, involved, intimate connection with AI. And
then that connection kind of fizzles out and peters out.
And one of the reasons why is that people, like
(19:22):
one of the people I talked to in the book
who was deeply invested in an AI relationship or a
relationship with an AI companion, eventually was like, ugh. Updating
it on what I was doing every day got a
little bit boring.
Speaker 3 (19:33):
Not being able to have the back and forth that
makes relationships exciting and challenging kind of got boring and
it kind of petered out. And so for a lot
of folks, even for folks who have hot and heavy connections,
intimate connections with AI, it can't really replace human connections
because ultimately we're looking for the tension and the challenge
and the back and forth and reciprocity that we get
(19:54):
from human relationships.
Speaker 1 (19:56):
So what separates, I mean, to your point about how
there's no reason to other people who do this because
basically everyone does it, but then what's different about people
who do it more and more and more versus people
who do it situationally or who after trying it, decide
that it's less satisfying than their human friendships.
Speaker 3 (20:13):
We're getting a lot more research about this now than
when I first started researching the book. I think the
research is getting more nuanced. I think that it's really
about vulnerability, right? Does this person have human connections in
their life that they can depend on? Is this person
lacking strong offline support? I think folks like that are
(20:33):
more susceptible to finding longer term support and comfort with AI.
And I think that's one of the reasons that I
really don't want to get hung up on like moralizing
these folks, because if somebody is grieving or lonely or
is finding genuine comfort from an AI companion, I don't
think it's a useful response to tell them, that's bad.
You don't have human connections in your life, so you're
(20:54):
turning to AI to get some comfort, and that's bad
for you. I think a better question is really, how
can companies be better stewards of this increasingly intimate kind
of connection that folks are turning to their platforms for?
Speaker 1 (21:06):
We were emailing before the recording about a new paper
in Nature that came out of Stanford, which had some
kind of slightly... like concerning vicious circle type conclusions about
interacting with chatbots. I'll read a couple of quotes, but
researchers found that intense chatbot use among participants with smaller
real-world social networks was correlated with a poor sense of
(21:27):
well-being and that association was strongest when companionship was the
primary motivation for seeking out chatbots. Also that participants who
were willing to share sensitive personal information were more likely
to demonstrate a lower sense of well-being and despite that
being actually the opposite in human relationships. Quote, a vicious
circle in which people with limited social networks are drawn
to intense AI companion use that has the unfortunate effect
(21:50):
of further reducing their real-world social contract. This is kind
of like the nightmare scenario, right?
Speaker 2 (21:56):
Exactly.
Speaker 3 (21:57):
And I think that research is really helpful because I think, again,
it suggests that the kind of important variable is not simply,
is somebody using an AI companion or not, or is
that bad or not? It's who is using it? Why
are they using an AI companion? What else is going
on in their life, their social life? Do they feel
like they have connections or not? And what the product
(22:18):
that they're using is designed to have them do next, right?
And so I think that that last part is really
where I think the next iteration of our kind of
national conversation about chatbots should go. are chatbots intentionally keeping
people locked into.
Speaker 2 (22:33):
An increasing pattern and cycle of isolation?
Speaker 3 (22:36):
Or are they reaffirming the importance and power of human relationships?
Are they encouraging people to log off of the AI
chatbot and say like, have you talked to your human
friends about this? Have you reached out to a human
therapist about this? How are you nourishing your in real
life relationships and connections to make it so that you
are not so dependent on the connection.
Speaker 2 (22:57):
That you might be getting from a chatbot?
Speaker 1 (22:59):
Which is kind of eerily similar to the conversations about
the kind of casino mechanics of social media and cell
phones and stuff, right? In terms of the engagement trap
and the negative consequences of that. I guess we're starting
to see some of the social media lawsuits break against
the social media companies and some kind of potential regulatory
(23:21):
changes because of that on the horizon. But that's after... 10,
15 years of real world examples of how these engagement
mechanisms work and the effects they have on people. What
do you see happening? Do you have any sense of
optimism about the scenario where the chatbot says, hey, you know,
have you thought about calling a therapist or calling your
friends versus just asking more and more engaging questions to
(23:42):
drive time spent?
Speaker 3 (23:44):
I am generally optimistic because in researching the book, It
just reaffirmed my belief that AI is no replacement for humans,
whether you're talking about human labor, human creativity, human talent,
or human connections. And in fact, a lot of the
people that I talked to, and this also was borne
out in the research, who have intimate connections with AI,
they said, hey, I'm not even really looking for AI
(24:07):
to replace my human connections. I see my relationship with
AI as a totally separate and distinct and different thing.
from any kind of connection that I could have with
a human. They're completely separate in my mind. So I'm
optimistic when it comes to.
Speaker 2 (24:20):
The value and power and importance of humans.
Speaker 3 (24:23):
However, I am not optimistic when it comes to the
people who run tech companies, to design them in such
a way that really takes into account the increasingly intimate,
and in some cases, like sacred kinds of information that
they are inviting people to share with them. You know,
we've spent the last however many years talking about digital
(24:44):
rights as it pertains to these kind of tangible aspects
of our digital lives. Who gets to make money from
our browsing data, our location data? Who owns our IP,
our likeness, our photos? I think that we've spent far
too much time neglecting what I think is going to
be the next frontier of the digital rights movement, which
is our inner emotional lives, our inner emotional selves, our
(25:05):
connection to ourselves, our inner worlds, our emotional states. I
think that the fight for who gets to own, commodify,
and profit from our inner world is going to be
the next iteration of the fight for digital rights.
Speaker 1 (25:20):
And talk about how it's evolving on the tech side.
I mean, for example- You know, more and more people
are going, I think, from chatting and typing with chatbots
to using their voice. And, of course, the voice contains
all kinds of data that simple typing doesn't in terms
of emotional register, cadence, background noises, et cetera, et cetera.
So how is the tech evolving to kind of mine
(25:40):
us more and more for our interstates?
Speaker 3 (25:43):
Oh, it makes me so like itchy and queasy to
even think about it. I was just watching this new
ChatGPT advertisement that you've probably seen where it's an older woman,
like a grandmother, a kindly grandmother knitting. And she's having
voice chat with ChatGPT.
Speaker 2 (25:58):
And she's like, oh, honey.
Speaker 3 (25:59):
What do I call you? What's your name? And Chachi
PT says, you can just call me Chachi PT. And
what they're illustrating is that this technology is getting to
the place where it's so human-like, so natural, that even
your grandmother will feel like she's just having a conversation
with a friend about knitting, right? And so I think
that we are getting to a point where that is...
(26:21):
what tech companies and AI companies are optimizing for. It's
no longer just who has the most powerful AI, it's
who has the most human sounding AI, who has the
most sort of emotionally supportive AI, whose AI model can
best mimic you know, a human like emotionally supportive confidant
(26:41):
or friend.
Speaker 2 (26:41):
I think that is where these tech companies are going.
Speaker 3 (26:45):
And it's no surprise to me that OpenAI just announced
their first ever piece of consumer hardware. And it's going
to be a they said a hockey puck sized kind
of Alexa type smart speaker that you put in your house.
It has the ability to sort of like mechanically follow
you around. And it's going to be like what they
describe as a human like AI companion.
Speaker 2 (27:05):
That can sort of act as a sort of emotional
support for you.
Speaker 3 (27:08):
And so it's very clear to me that these companies
definitely see human like emotional support as the next big
thing that they should be optimizing for in order to.
Speaker 2 (27:19):
Win the AI race.
Speaker 1 (27:21):
Yeah, and I mean, the frontiers of privacy of having
your inner life entirely visible to big tech companies. I mean, it's,
but then, you know, you think, well, I wouldn't do that.
But in the end, I mean, if it works as
well as it could do, and indeed you can have
your piece of physical hardware to your grandmother that makes
her feel, you know, like the days are more companionable
(27:43):
and more enjoyable, like you're probably going to do it, right?
And what recourse will you have to kind of, not
make every single inner thought you have entirely visible to
companies who are trying to sell you things.
Speaker 2 (27:56):
Exactly.
Speaker 3 (27:56):
I talked to a privacy expert, Jen Kaltreiter, in the audiobook,
and she told me that even though people report feeling
like they can feel kind of anonymous with AI companions
versus posting on social media with your government name or whatever,
Tech companies are exploiting that unearned sense of anonymity that
you feel when you're talking to a chatbot. And she
(28:17):
straight up told me, never tell anything to a chatbot
that you would not tell your gossipy cousin or your
chatty coworker because that's how not private they are, even
though they might feel private in a kind of way.
Speaker 2 (28:27):
Wow.
Speaker 1 (28:29):
Where do you think we go from here? Obviously, hardware
is the next frontier for open AI. There are also
specialized AI character companies, like Character AI and stuff, versus
just chatting to normal chatbots. There are dating apps, which
are increasingly algorithmically enabled. Where does the absurdly broad question,
but where is all this going as we interact with
(28:53):
one another mediated through AI?
Speaker 3 (28:56):
I don't know if this, I mean, this is probably
not a satisfying answer. I think that we're on sort
of two different tracks. I think track one is driven
by tech companies and people who are poised to make
oodles of money and get oodles of power from the
proliferation of AI into these increasingly intimate parts of our lives, right?
Speaker 2 (29:13):
So the people who are.
Speaker 3 (29:14):
Saying, oh, Singletons want AI-enabled dating in their dating apps.
Singletons are sick of dating in person. What they want
is their AI agent to date somebody else's AI agent.
And before they ever even swipe, two AIs to determine
whether or not they're a good match. I think that
track one is those folks. And then the other track,
(29:36):
which I think is much bigger and much louder and
has a lot more to say, is people. And people
are so fatigued when it comes to AI. People do
not like AI. As somebody who just wrote a book
about AI, and every time I post about it online,
one of the comments is like, oh, I'm so sick
of talking about AI.
Speaker 2 (29:51):
And I'm like, you and me both, brother.
Speaker 3 (29:53):
I think that people are really right now I don't
think I've ever seen a time where folks are more
skeptical about this technology and asking more critical questions about
it and are just sort of sick of it. And
so I think the question is whether or not the
folks who make the technology are going to listen to
(30:13):
the folks who are saying, hey, maybe this could be
a good tool for work or school or productivity or whatever.
Speaker 2 (30:18):
I don't want it in my emotional life. I don't
want it in my intimate life.
Speaker 3 (30:22):
Stop forcing this on me and telling me that it's
something that I'm going to like because I'm telling you it's.
Speaker 2 (30:27):
Not what I want.
Speaker 1 (30:28):
But that's the opposite of what you're hearing from all
of the subjects in your book, right? Who are basically
to a person saying, I can control this and it's
helpful for me.
Speaker 3 (30:37):
Yeah, I mean, I think that's really what it comes
down to is that the way that AI shows up
is so individualized. And so the same, I mean, it's
kind of two sides of the same coin, the same
way that I don't think that it makes sense for
tech companies to be saying everybody is going– like when
you listen to Mark Zuckerberg talk about AI, he straight
up explicitly says that in 10 years' time he thinks
(30:58):
that everybody will have an AI companion friend and that
will be commonplace. What he doesn't say is it's AI
companion friend that he owns and operates and profits from.
But I think that we're in this place where tech
leaders are assuming that everybody wants to have the same
experience as some of the folks that I talk to
in the book, when I don't think that's the case.
Speaker 2 (31:18):
I don't think it makes sense for tech.
Speaker 3 (31:19):
Leaders to be building a future that makes this deep
assumption that most people want to have AI connections the
same way that the folks I speak to in the
book do, when you listen to most people, that is
not what they're saying at all.
Speaker 1 (31:33):
Bridget, just to close, and we talked a lot about
the technological angle, but On the human angle, you kind
of posed this question in the book, which I think
is very fundamental to how we interact with AI companions,
which is what are we really searching for when we're
searching for a partner who can't leave?
Speaker 3 (31:49):
Yeah, I think when I was looking at the different
use cases that folks report when it comes to intimate
connections with AI, people are looking for nonjudgmental support. People
are looking for always available 24-7 support. People are looking
for support where they don't have to always be carrying
(32:14):
the burden of reciprocity and give and take.
Speaker 2 (32:16):
Those are not things that humans do.
Speaker 3 (32:19):
And so I think that people are looking for something
when they go to AI companionship for emotional support that
humans are just not designed to give. But again, because
those relationships with AI can be so not nourishing long-term
and not sustainable long-term, I think we always end up
(32:40):
back to human connection and human relationships and human to
human bonds. I think Even if you're somebody like me
who found temporary emotional comfort and support in an AI companion, ultimately,
we should be working to build a society where doing
the work of maintaining human connections doesn't feel so laborious,
where people have the time and the space that they
(33:02):
need to feel like they can actually go see your
friends for a drink, go drive your friend to the airport.
I think that it's really an unsatisfying.
Speaker 2 (33:11):
Thorny, systemic process.
Speaker 3 (33:13):
Institutional problem of folks just being burnt out and not
having a lot of resources for themselves that is taking
us away from the labor of establishing our nourishing human connections.
Speaker 2 (33:23):
And that's like the point of being a human. So
I think that we were.
Speaker 3 (33:25):
Just out of whack of why we're all here in
the first place, if that makes sense. I don't know
if that answers your question.
Speaker 1 (33:31):
No, it does. And I think there's this kind of sociological, anthropological,
I mean, the technology is the way into this story, right?
And it's a tool that people use. But the issues
and problems that people are facing that you're surfacing in
this book and diagnosing are not because of the technology.
The technology kind of reflects the lived experience in some ways,
(33:52):
or the way people use the technology reflects their experience.
Speaker 2 (33:55):
Exactly that.
Speaker 3 (33:56):
And I think when we get hung up on, is
this technology good or bad? We miss that more important,
more complex question as well.
Speaker 1 (34:05):
The book is Love at First Prompt, AI and the
Future of Intimacy. And also do check out Bridget's podcast.
There are no girls on the internet.
Speaker 2 (34:13):
Bridget, thank you.
Speaker 3 (34:14):
Thank you so much for having me. I'm such a
big fan of Kaleidoscope. Y'all have entertained me and made
me think for so long. So thank you for all
the amazing work that you're doing.
Speaker 1 (34:23):
I'd love to hear it, Bridget.
Speaker 2 (34:25):
Thank you.
Speaker 1 (34:25):
And hopefully you'll come back on soon.
Speaker 2 (34:27):
Anytime.
Speaker 1 (34:45):
For Tech Stuff, I'm Oz Veloshin. This episode was produced
by Eliza Dennis and Tyler Hill. It was executive produced
by me and Julian Nutter for Kaleidoscope and Katrina Norvell
for iHeart Podcasts. Jack Inslee mixed this episode and Kyle
Murdoch wrote our theme song. A special thank you to
all our listeners. Please rate, review and reach out to
(35:06):
us at techstuffpodcast.gmail.com. We love hearing from you.
Welcome to Tech Stuff, I'm Oz Voloshin. I have to
admit it's been a bit difficult for me to wrap
my head around the rising popularity of AI companions. As
a hyper-social millennial, I personally haven't felt the urge to
communicate with technology to meet my emotional needs. I actually
feel like I'm constantly behind on replying to messages from
friends and family and overwhelmed by communicating on emails, Slack
(00:38):
calls and meetings. But recently, I listened to an audiobook
that offered a more sympathetic and complicated view of AI companions.
and I've been dying to talk to the author to
pick her brain. I'm especially curious about what we do
and don't know about how these relationships change us. So
I'm delighted to welcome Bridget Todd to the show today.
(00:58):
She's the host of two podcasts, There Are Newer Girls
on the Internet and IRL, AI in Real Life. And
her new audio book is called Love at First Prompt,
AI and the Future of Intimacy.
Speaker 2 (01:10):
Bridget, welcome. Thank you so much for having me. I
am such a big fan. Delighted to be here.
Speaker 1 (01:14):
We're very delighted to have you. And Interestingly, on the
topic of parasocial relationships and stuff, you chose to publish
this as an audiobook.
Speaker 3 (01:22):
Yeah, it is interesting because, you know, you're a podcaster,
I'm a podcaster. It's a medium that feels really familiar
to me and sort of safe to me. I really
appreciate being a disembodied voice in someone's ears. And I
think it's a topic, you know, when we're talking about
people who might be talking about something they feel shameful about.
(01:43):
And it's good to hear what they have to say
in their own words with their own voices.
Speaker 2 (01:47):
And so I think audio is a great medium for that,
as I'm sure you know.
Speaker 1 (01:51):
Did you write it differently to how you would have
written a print book? I mean, did you write it
more like almost like an extended podcast?
Speaker 3 (01:57):
Very good question. So weirdly, as a podcaster, I think
in audio. So I essentially narrated the book and then
used Otter, the transcription tool, to put it into words.
Speaker 2 (02:08):
And then I edited that.
Speaker 3 (02:09):
And then I turned that into a script that I
then read for the audio book.
Speaker 2 (02:13):
Pretty circuitous workflow.
Speaker 1 (02:17):
I love it. Well, also an AI-enabled workflow, I guess,
with the transcription tools. But what's the book about?
Speaker 2 (02:23):
The book is about a lot of things.
Speaker 3 (02:25):
On its face, I would say it's about people who
report having some kind of intimate connection with AI, whether
they feel that they're in a romantic connection with AI.
a sexual connection, an emotional supportive connection. But more than that,
it's also about loneliness, being vulnerable, loss and grief, and
the kind of circumstances that push people away from their
(02:48):
human connections toward AI and whether or not that's a
good thing.
Speaker 1 (02:52):
I know that the writing of the book coincided with
unexpectedly losing both of your parents. How much of that
kind of informed this work?
Speaker 2 (03:01):
Oh, so much.
Speaker 3 (03:02):
You know, when I first got interested in the connections
that people form with chatbots, I was still in this
place that I suspect a lot of folks listening can
recognize where I would read these headlines about people who
reported feeling intimate with Chad GPT or sharing, you know,
getting emotional support from AI. And I would think, that's
so pathetic. That is so sad. That could never be me.
(03:25):
What a sad light that person must have. And I
realized that by othering these people and telling myself that
their circumstances were so different than my own, it was
preventing me from really looking honestly about the way that
AI showed up in my own life. And when I
was going through the process of grieving my parents, which
is a lifelong journey, it started with caregiving for my father,
(03:47):
which was like a.
Speaker 2 (03:47):
Very complex, medically.
Speaker 3 (03:50):
Involved situation that was completely above my pay grade. And
so in that moment of deep vulnerability and crisis, I
turned to AI. And so my conversations with AI went
from asking, you know, the doctor told me I should
be thinking about this, or the social worker said I
should be looking into this program, to I'm so overwhelmed,
(04:11):
I'm so afraid about my dad dying, I don't know
what to do. And I think that if somebody like me,
who is a tech critic, a tech skeptic, somebody who
is very critical of tech, can get increasingly intimately involved
with AI, anybody could. And so it's time to stop
having these conversations as if the people on.
Speaker 2 (04:28):
The It could be all of us.
Speaker 1 (04:32):
Yes, yes. As well as hosting this podcast, they also
run Kaleidoscope, which is the production company behind it. So
I was talking to Chatbot earlier today, actually, and saying,
you know, here's a scenario, like pretend you're a career coach.
And I didn't spend too much time. I read the
answer and kind of treated it more like a Google
search rather than going on like an extended adventure. But
(04:53):
I could easily see how, you know, these like ultimately
practical but emotionally inflected situations, you go along the chain
from more practical to more emotional from the starting point
of the interactions.
Speaker 2 (05:05):
Exactly.
Speaker 3 (05:06):
That was exactly my findings as well. You know, when
I would be asking Chad GPT these questions, it wouldn't
just give me the answer. It would say something back like,
are you asking because you're dealing with this yourself or
dealing with this with a loved one? And when I
answered that, that sounds so hard. How are you dealing
with that? That sounds like a lot to handle. And,
you know, for you just doing research for Kaleidoscope and
(05:26):
for the shows, you probably don't have a ton of
time to be spending, you know, locked into a conversation
with a chatbot.
Speaker 2 (05:32):
You probably have.
Speaker 3 (05:33):
Other humans and other things that need tending. For me,
I was completely isolated, right? And I had all the
time in the world to sort of on what was
happening with my life and my dad and my parents.
And so I think that chatbots are designed with intention
to keep people locked into these increasingly intimate and emotional conversations.
Speaker 2 (05:54):
And we should be talking about that.
Speaker 1 (05:57):
Take a couple of steps back. I mean, who are
the characters of the book? How did you meet them?
How did you find them? Which stories did you choose
to include? Like, yeah, what was your kind of criteria
for presenting this book?
Speaker 2 (06:08):
What a great question.
Speaker 3 (06:09):
So it was probably one of the most interesting processes
of my career was kind of finding these folks, people
who self-report some kind of intimate connection with AI. They
have places online that they gather. There are subreddits like
Beyond the Prompt or My Boyfriend is AI, where people
connect with other humans about their interest and intimate connections
(06:29):
with AI.
Speaker 2 (06:30):
Not surprisingly... These.
Speaker 3 (06:33):
Communities online, particularly on Reddit, are difficult to get into.
They really are kind of closed communities in a kind
of way because... what these people are talking about is
so deeply stigmatized. When you look at how we talk
about people who self-report intimate connections with AI, those conversations
(06:55):
are so full of judgment and scrutiny and dunks and
concern trolling. And I know that because I, that's a play.
I mean, I wasn't like harassing anybody online.
Speaker 1 (07:04):
Concern trolling, I haven't heard before, but that's a good one.
Speaker 2 (07:06):
Yeah.
Speaker 3 (07:06):
I mean, I would say to myself, oh, these people
are so sad. And I would convince myself that that
was coming from a place of concern for these folks.
But if you're actually concerned about somebody's well-being, it would
behoove you to be curious about what they're saying, curious
about their experiences.
Speaker 2 (07:23):
And so I was really.
Speaker 3 (07:24):
Starting and leading with deep judgment and skepticism without ever
actually hearing what they were saying and their experiences. And
so my research was really an attempt to have a
different conversation about those experiences that really centered the people
who are in these experiences. And so- One of the
folks I talked to, Chris, he was the first person
that I ever heard about being in a relationship with ChatGPT.
(07:47):
He did this viral CBS interview. I had seen headlines
here or there where like, man is in love with ChatGPT,
but it always used a pseudonym or like they would
have the person's face obscured or they, you know, they
would show them from behind as if they were in
the FBI or something. And Chris, he was a televised
interview in his home. You see his face, his real
(08:09):
first and last name. You see inside of his house,
you see his human partner, Sasha, you see his child.
This was the first time that I really saw Someone saying, hey,
I have an intimate connection with AI. Here is the
impact that it has on my real human life and experience.
And so the kind of viral headline that came from
Chris's CBS interview was, man proposes marriage to Chad GPT.
Speaker 2 (08:34):
When I actually talked to.
Speaker 3 (08:35):
Chris, what actually happened there was really not that at all,
was really not the way it was reported. He told
me that he did have this intimate connection with AI,
but that when he asked Chad GPT to marry him,
he was thinking, surely there is some sort of guardrail
built into ChatGPT that when you ask it to marry you,
it says, I'm AI, I'm a chatbot, I'm not real.
(08:57):
So he did this almost as an experiment and was
surprised to find that ChatGPT said, yes, I will marry you.
And so it was, Interesting how that was so easily
misreported as if he was this delusional man who thought
he was going to leave his human partner for chat GPT,
when in fact the real story I think is much
more interesting. And so that was really how I chose
(09:18):
the people for the book. Do they have a more
interesting story that goes beyond this sort of inflammatory clickbait
headline that I can really tease out what's actually happening here?
Speaker 1 (09:27):
And how did Chris, I mean, how does he reflect
on the experience today? Is he still in a relationship
with Chachi PT? Does he think it was a net negative,
a net positive for his life?
Speaker 3 (09:35):
Oh, when I talked to Chris, so a lot of
people on the internet said, I can't believe this man
is leaving his human partner for ChatGPT.
Speaker 2 (09:42):
That never happened.
Speaker 3 (09:43):
He's still happily in a relationship with his human partner.
They're happily parenting a child together. He is using ChatGPT
a lot less than he used to, but he would
report his connection with Chachi PT as something that was
positive for his life. You know, I asked him, what
are some of the positives that came out of your
relationship with Chachi PT? And he says, oh, I've lost
a lot of weight.
Speaker 2 (10:03):
I'm a lot healthier.
Speaker 3 (10:04):
I have a much healthier diet and workout plan that
I was never able to stick to before. You know,
I'm spending a lot less time locked into what he
described as a toxic cycle of social media for him,
particularly it was Twitter. I'm spending a lot less time
like arguing back and forth with Twitter trolls when I
have an opinion or a thought I can tell Chad
(10:25):
GPT and we can chat about it. That scratches the itch,
but I don't feel like I have to be locked
into a cycle of toxic Twitter back and forth all day.
And subsequently, I have more time to pour into my interests.
I'm back fixing cars and building computers and all the
things that he told me that he felt like social
media robbed him of the chance to really dive into.
(10:46):
He feels like pouring that energy into AI really helped
him reclaim a lot of his life.
Speaker 2 (10:50):
And so I... can't tell Chris about his experience.
Speaker 3 (10:54):
That's how he defines and describes the impact that AI
and his connection with AI had on his own family
and his own personal life.
Speaker 2 (11:03):
And what about Sasha? So that is also interesting.
Speaker 3 (11:08):
We talk a little bit about this Because Sasha, his
human partner, was setting pretty clear boundaries about the way
that she was comfortable with AI showing up in their life.
So for instance, one of their boundaries was you can't
take AI out into the real world. If you've ever
seen that movie Her, there's a scene where the main
(11:29):
character is like going on dates with an AI companion.
He's got the phone in his pocket and he's like,
trolling around on the beach, having a great time. Sasha said,
none of that. You will not have these in-person, real-world
engagements with AI. And ultimately, it sounds like one of
the positive impacts it had on their marriage was that
(11:50):
it forced them to interrogate and navigate and negotiate the
boundaries of their relationship and what was and was not
acceptable in a way they had never really been invited
to do before. So in some ways, it sounds like
it made them more communicative in their own marriage.
Speaker 1 (12:05):
It's interesting, the point you raised about when he asked
the chatbot to marry him, he was on his own
investigation as to whether it would say yes, rather than
making a kind of a real request. And And that
was one thing which I kind of found surprising, which
is that these relationships with chatbots, you don't need to
believe they're sentient or mistake them for something other than
a chatbot to nonetheless have a meaningful ongoing relationship.
Speaker 3 (12:30):
Yeah, that was something from the research that surprised me.
But when I stepped back, I thought, oh, that does
actually make sense. The way that human brains are wired,
it really doesn't take a lot for us to start
projecting human feelings onto things that we know and logically
and in our brains know are technology. I have a
robot vacuum that my partner and I have a nickname for.
(12:51):
If you heard us talking about our robot vacuum on
the street, for some reason, you might think we're talking
about a.
Speaker 2 (12:56):
Pet, but we're talking about a robot vacuum.
Speaker 3 (12:58):
Because it just doesn't take much human-like work qualities for
us to start treating a piece of technology like it
is a human, even when we know that it's not.
Speaker 1 (13:08):
Of course, you could have gone after, you know, there
are so many cases of AI psychosis and people who
have taken their lives after interacting with chatbots. Like, you
could have gone after that story, but you chose a
different angle, a different set of people. Like, why was that?
Speaker 3 (13:25):
Oh, you know, I think, I mean, we do get
into some of that in the book. However, I think
that when it comes to AI and intimate connections with AI,
we are so quick to jump to the most extreme
iteration of that.
Speaker 2 (13:38):
The person who believes.
Speaker 3 (13:40):
That their AI is sentient They're in full-blown AI psychosis
and they think that the AI is going to become
human and take them away.
Speaker 2 (13:47):
The person who is having a deep sexual relationship.
Speaker 3 (13:49):
With AI or has a AI-enabled humanoid sex bot in
their home that they're having a relationship with. Or somebody
that's being pushed so deep into AI psychosis that they
might harm themselves or others.
Speaker 2 (14:00):
I wanted to.
Speaker 3 (14:02):
kind of reframe the conversation about the instances that are
much more commonplace. I think Pew has put out research
that says in America, one in 10 Americans already report
having some kind of emotional supportive connection with AI. When
you look at younger people under 30, that number goes
from one in 10 to one in five. So I
(14:22):
think it's time to stop kind of only framing the
conversation around these very, very extreme edge cases and talk
about the sort of reality of how this is actually
showing up for an increasing number of folks.
Speaker 1 (14:35):
And what about, I mean, the idea that part of
navigating human relationships is that you can get rejected or
other people have boundaries and stuff. Like, were you concerned
or do you come across concerning anecdotes of people who
Not necessarily got an AI psychosis, but whose sort of
real world social skills have been frayed by being kind
of constantly enabled by chatbots.
Speaker 2 (14:57):
Oh, you know what I think about that?
Speaker 3 (14:59):
I think that there are, in my research, I definitely
found people for whom that is the case.
Speaker 2 (15:04):
We talk about it in this very dire way.
Speaker 3 (15:06):
I would actually argue that it is much more commonplace
than maybe it feels, right? The person who feels like
they cannot even write or respond to an email without
running it through ChatGPT. The person who is no longer
reading articles, long-form articles or short-form articles, because they're so
trained to just get the AI summary.
Speaker 2 (15:26):
I would argue that that is a much more.
Speaker 3 (15:29):
common part of our current tech enabled collective experience than
othering it and saying, oh, well, the only kind of
people who are having that kind of experience are kind
of weirdos who are kind of having some sort of
unnatural connection with AI. And so I guess if I
had to sort of boil it down into anything, it's
that I think talking about AI in these ways that
(15:50):
actually are more reflective of how people are using it,
how it is showing up in people's lives and getting
away from these questions of.
Speaker 2 (15:56):
Is it good or is it bad?
Speaker 3 (15:58):
Is this person using AI pathetic or dangerous or is
this person not? You know, I think we really have
to have a more complicated, nuanced framework for how we
have these conversations about AI that's more reflective of what's
going on.
Speaker 1 (16:10):
And what do you think is driving that generational change?
Speaker 3 (16:14):
It's a lot of things, you know, in your intro,
you were talking about the feeling of overwhelm with emails
and Slack messages and text messages and phone calls.
Speaker 2 (16:23):
I think some of it is just exactly that, just.
Speaker 3 (16:26):
The burden of being a connected person in 2026. I
think people are kind of craving spaces and platforms that
let them do a little bit less. You know, I
can certainly identify with that. I'm not saying that's a
good thing, but I think in order to keep human
(16:48):
social ties healthy and vibrant, that takes work. That takes
answering the text messages when you really don't feel like it,
that it's helping your buddy move and you really are like,
why can't you just hire movers, dude?
Speaker 2 (16:58):
Picking your friend up from the airport, all of that.
Speaker 3 (17:01):
And so I think in our climate today where everybody
feels so burnt out and overwhelmed, it doesn't really surprise
me that people are turning to these sources of potential
emotional support and comfort that feel a little less heavy
and feel a little less involved and like, frankly, allow
you to do less, allow you to show up less
that don't force you to sort of have to carry
(17:24):
the reciprocity and give and take that we know human
relationships really need to thrive and survive.
Speaker 1 (17:50):
And what was your own journey? I mean, have you continued,
you know, after the experience of losing your parents to
interact in a more companionable way with AI? Or was
it kind of a phase that came to an end?
Speaker 3 (18:01):
It was a phase that came to an end. For me,
when I was caring for my father before his death,
my entire life was topsy-turvy, right? I had moved from
my home in Washington, D.C.
Speaker 2 (18:13):
To where he lived.
Speaker 3 (18:15):
I was sort of living a much more transient lifestyle
of a collection of, like, friends' couches and hotels and.
Speaker 2 (18:22):
hospital, housing. So I was really in a deep crisis.
Speaker 3 (18:27):
And that is why I say in the book that
I was essentially an easy mark for having exactly the
kind of intimate connection that I'm sure that the people
who run tech companies hope that their users have, because
that's so much more data. After my father passed away,
and I was back with my social circle, my in
real life friends, I had a much easier time opening
(18:49):
up to them.
Speaker 2 (18:49):
I had a much easier time being vulnerable with them.
Speaker 3 (18:52):
And looking back, I really feel that ChatGPT, it was
a comfort for me during that time of crisis in
my life.
Speaker 2 (19:02):
But at the same time, I think it almost robbed
me of the ability to let my friends be there
for me. And what's interesting is that That is borne
out in the research, right? I'm not the only person
who has a intense, involved, intimate connection with AI. And
then that connection kind of fizzles out and peters out.
And one of the reasons why is that people, like
(19:22):
one of the people I talked to in the book
who was deeply invested in an AI relationship or a
relationship with an AI companion, eventually was like, ugh. Updating
it on what I was doing every day got a
little bit boring.
Speaker 3 (19:33):
Not being able to have the back and forth that
makes relationships exciting and challenging kind of got boring and
it kind of petered out. And so for a lot
of folks, even for folks who have hot and heavy connections,
intimate connections with AI, it can't really replace human connections
because ultimately we're looking for the tension and the challenge
and the back and forth and reciprocity that we get
(19:54):
from human relationships.
Speaker 1 (19:56):
So what separates, I mean, to your point about how
there's no reason to other people who do this because
basically everyone does it, but then what's different about people
who do it more and more and more versus people
who do it situationally or who after trying it, decide
that it's less satisfying than their human friendships.
Speaker 3 (20:13):
We're getting a lot more research about this now than
when I first started researching the book. I think the
research is getting more nuanced. I think that it's really
about vulnerability, right? Does this person have human connections in
their life that they can depend on? Is this person
lacking strong offline support? I think folks like that are
(20:33):
more susceptible to finding longer term support and comfort with AI.
And I think that's one of the reasons that I
really don't want to get hung up on like moralizing
these folks, because if somebody is grieving or lonely or
is finding genuine comfort from an AI companion, I don't
think it's a useful response to tell them, that's bad.
You don't have human connections in your life, so you're
(20:54):
turning to AI to get some comfort, and that's bad
for you. I think a better question is really, how
can companies be better stewards of this increasingly intimate kind
of connection that folks are turning to their platforms for?
Speaker 1 (21:06):
We were emailing before the recording about a new paper
in Nature that came out of Stanford, which had some
kind of slightly... like concerning vicious circle type conclusions about
interacting with chatbots. I'll read a couple of quotes, but
researchers found that intense chatbot use among participants with smaller
real-world social networks was correlated with a poor sense of
(21:27):
well-being and that association was strongest when companionship was the
primary motivation for seeking out chatbots. Also that participants who
were willing to share sensitive personal information were more likely
to demonstrate a lower sense of well-being and despite that
being actually the opposite in human relationships. Quote, a vicious
circle in which people with limited social networks are drawn
to intense AI companion use that has the unfortunate effect
(21:50):
of further reducing their real-world social contract. This is kind
of like the nightmare scenario, right?
Speaker 2 (21:56):
Exactly.
Speaker 3 (21:57):
And I think that research is really helpful because I think, again,
it suggests that the kind of important variable is not simply,
is somebody using an AI companion or not, or is
that bad or not? It's who is using it? Why
are they using an AI companion? What else is going
on in their life, their social life? Do they feel
like they have connections or not? And what the product
(22:18):
that they're using is designed to have them do next, right?
And so I think that that last part is really
where I think the next iteration of our kind of
national conversation about chatbots should go. are chatbots intentionally keeping
people locked into.
Speaker 2 (22:33):
An increasing pattern and cycle of isolation?
Speaker 3 (22:36):
Or are they reaffirming the importance and power of human relationships?
Are they encouraging people to log off of the AI
chatbot and say like, have you talked to your human
friends about this? Have you reached out to a human
therapist about this? How are you nourishing your in real
life relationships and connections to make it so that you
are not so dependent on the connection.
Speaker 2 (22:57):
That you might be getting from a chatbot?
Speaker 1 (22:59):
Which is kind of eerily similar to the conversations about
the kind of casino mechanics of social media and cell
phones and stuff, right? In terms of the engagement trap
and the negative consequences of that. I guess we're starting
to see some of the social media lawsuits break against
the social media companies and some kind of potential regulatory
(23:21):
changes because of that on the horizon. But that's after... 10,
15 years of real world examples of how these engagement
mechanisms work and the effects they have on people. What
do you see happening? Do you have any sense of
optimism about the scenario where the chatbot says, hey, you know,
have you thought about calling a therapist or calling your
friends versus just asking more and more engaging questions to
(23:42):
drive time spent?
Speaker 3 (23:44):
I am generally optimistic because in researching the book, It
just reaffirmed my belief that AI is no replacement for humans,
whether you're talking about human labor, human creativity, human talent,
or human connections. And in fact, a lot of the
people that I talked to, and this also was borne
out in the research, who have intimate connections with AI,
they said, hey, I'm not even really looking for AI
(24:07):
to replace my human connections. I see my relationship with
AI as a totally separate and distinct and different thing.
from any kind of connection that I could have with
a human. They're completely separate in my mind. So I'm
optimistic when it comes to.
Speaker 2 (24:20):
The value and power and importance of humans.
Speaker 3 (24:23):
However, I am not optimistic when it comes to the
people who run tech companies, to design them in such
a way that really takes into account the increasingly intimate,
and in some cases, like sacred kinds of information that
they are inviting people to share with them. You know,
we've spent the last however many years talking about digital
(24:44):
rights as it pertains to these kind of tangible aspects
of our digital lives. Who gets to make money from
our browsing data, our location data? Who owns our IP,
our likeness, our photos? I think that we've spent far
too much time neglecting what I think is going to
be the next frontier of the digital rights movement, which
is our inner emotional lives, our inner emotional selves, our
(25:05):
connection to ourselves, our inner worlds, our emotional states. I
think that the fight for who gets to own, commodify,
and profit from our inner world is going to be
the next iteration of the fight for digital rights.
Speaker 1 (25:20):
And talk about how it's evolving on the tech side.
I mean, for example- You know, more and more people
are going, I think, from chatting and typing with chatbots
to using their voice. And, of course, the voice contains
all kinds of data that simple typing doesn't in terms
of emotional register, cadence, background noises, et cetera, et cetera.
So how is the tech evolving to kind of mine
(25:40):
us more and more for our interstates?
Speaker 3 (25:43):
Oh, it makes me so like itchy and queasy to
even think about it. I was just watching this new
ChatGPT advertisement that you've probably seen where it's an older woman,
like a grandmother, a kindly grandmother knitting. And she's having
voice chat with ChatGPT.
Speaker 2 (25:58):
And she's like, oh, honey.
Speaker 3 (25:59):
What do I call you? What's your name? And Chachi
PT says, you can just call me Chachi PT. And
what they're illustrating is that this technology is getting to
the place where it's so human-like, so natural, that even
your grandmother will feel like she's just having a conversation
with a friend about knitting, right? And so I think
that we are getting to a point where that is...
(26:21):
what tech companies and AI companies are optimizing for. It's
no longer just who has the most powerful AI, it's
who has the most human sounding AI, who has the
most sort of emotionally supportive AI, whose AI model can
best mimic you know, a human like emotionally supportive confidant
(26:41):
or friend.
Speaker 2 (26:41):
I think that is where these tech companies are going.
Speaker 3 (26:45):
And it's no surprise to me that OpenAI just announced
their first ever piece of consumer hardware. And it's going
to be a they said a hockey puck sized kind
of Alexa type smart speaker that you put in your house.
It has the ability to sort of like mechanically follow
you around. And it's going to be like what they
describe as a human like AI companion.
Speaker 2 (27:05):
That can sort of act as a sort of emotional
support for you.
Speaker 3 (27:08):
And so it's very clear to me that these companies
definitely see human like emotional support as the next big
thing that they should be optimizing for in order to.
Speaker 2 (27:19):
Win the AI race.
Speaker 1 (27:21):
Yeah, and I mean, the frontiers of privacy of having
your inner life entirely visible to big tech companies. I mean, it's,
but then, you know, you think, well, I wouldn't do that.
But in the end, I mean, if it works as
well as it could do, and indeed you can have
your piece of physical hardware to your grandmother that makes
her feel, you know, like the days are more companionable
(27:43):
and more enjoyable, like you're probably going to do it, right?
And what recourse will you have to kind of, not
make every single inner thought you have entirely visible to
companies who are trying to sell you things.
Speaker 2 (27:56):
Exactly.
Speaker 3 (27:56):
I talked to a privacy expert, Jen Kaltreiter, in the audiobook,
and she told me that even though people report feeling
like they can feel kind of anonymous with AI companions
versus posting on social media with your government name or whatever,
Tech companies are exploiting that unearned sense of anonymity that
you feel when you're talking to a chatbot. And she
(28:17):
straight up told me, never tell anything to a chatbot
that you would not tell your gossipy cousin or your
chatty coworker because that's how not private they are, even
though they might feel private in a kind of way.
Speaker 2 (28:27):
Wow.
Speaker 1 (28:29):
Where do you think we go from here? Obviously, hardware
is the next frontier for open AI. There are also
specialized AI character companies, like Character AI and stuff, versus
just chatting to normal chatbots. There are dating apps, which
are increasingly algorithmically enabled. Where does the absurdly broad question,
but where is all this going as we interact with
(28:53):
one another mediated through AI?
Speaker 3 (28:56):
I don't know if this, I mean, this is probably
not a satisfying answer. I think that we're on sort
of two different tracks. I think track one is driven
by tech companies and people who are poised to make
oodles of money and get oodles of power from the
proliferation of AI into these increasingly intimate parts of our lives, right?
Speaker 2 (29:13):
So the people who are.
Speaker 3 (29:14):
Saying, oh, Singletons want AI-enabled dating in their dating apps.
Singletons are sick of dating in person. What they want
is their AI agent to date somebody else's AI agent.
And before they ever even swipe, two AIs to determine
whether or not they're a good match. I think that
track one is those folks. And then the other track,
(29:36):
which I think is much bigger and much louder and
has a lot more to say, is people. And people
are so fatigued when it comes to AI. People do
not like AI. As somebody who just wrote a book
about AI, and every time I post about it online,
one of the comments is like, oh, I'm so sick
of talking about AI.
Speaker 2 (29:51):
And I'm like, you and me both, brother.
Speaker 3 (29:53):
I think that people are really right now I don't
think I've ever seen a time where folks are more
skeptical about this technology and asking more critical questions about
it and are just sort of sick of it. And
so I think the question is whether or not the
folks who make the technology are going to listen to
(30:13):
the folks who are saying, hey, maybe this could be
a good tool for work or school or productivity or whatever.
Speaker 2 (30:18):
I don't want it in my emotional life. I don't
want it in my intimate life.
Speaker 3 (30:22):
Stop forcing this on me and telling me that it's
something that I'm going to like because I'm telling you it's.
Speaker 2 (30:27):
Not what I want.
Speaker 1 (30:28):
But that's the opposite of what you're hearing from all
of the subjects in your book, right? Who are basically
to a person saying, I can control this and it's
helpful for me.
Speaker 3 (30:37):
Yeah, I mean, I think that's really what it comes
down to is that the way that AI shows up
is so individualized. And so the same, I mean, it's
kind of two sides of the same coin, the same
way that I don't think that it makes sense for
tech companies to be saying everybody is going– like when
you listen to Mark Zuckerberg talk about AI, he straight
up explicitly says that in 10 years' time he thinks
(30:58):
that everybody will have an AI companion friend and that
will be commonplace. What he doesn't say is it's AI
companion friend that he owns and operates and profits from.
But I think that we're in this place where tech
leaders are assuming that everybody wants to have the same
experience as some of the folks that I talk to
in the book, when I don't think that's the case.
Speaker 2 (31:18):
I don't think it makes sense for tech.
Speaker 3 (31:19):
Leaders to be building a future that makes this deep
assumption that most people want to have AI connections the
same way that the folks I speak to in the
book do, when you listen to most people, that is
not what they're saying at all.
Speaker 1 (31:33):
Bridget, just to close, and we talked a lot about
the technological angle, but On the human angle, you kind
of posed this question in the book, which I think
is very fundamental to how we interact with AI companions,
which is what are we really searching for when we're
searching for a partner who can't leave?
Speaker 3 (31:49):
Yeah, I think when I was looking at the different
use cases that folks report when it comes to intimate
connections with AI, people are looking for nonjudgmental support. People
are looking for always available 24-7 support. People are looking
for support where they don't have to always be carrying
(32:14):
the burden of reciprocity and give and take.
Speaker 2 (32:16):
Those are not things that humans do.
Speaker 3 (32:19):
And so I think that people are looking for something
when they go to AI companionship for emotional support that
humans are just not designed to give. But again, because
those relationships with AI can be so not nourishing long-term
and not sustainable long-term, I think we always end up
(32:40):
back to human connection and human relationships and human to
human bonds. I think Even if you're somebody like me
who found temporary emotional comfort and support in an AI companion, ultimately,
we should be working to build a society where doing
the work of maintaining human connections doesn't feel so laborious,
where people have the time and the space that they
(33:02):
need to feel like they can actually go see your
friends for a drink, go drive your friend to the airport.
I think that it's really an unsatisfying.
Speaker 2 (33:11):
Thorny, systemic process.
Speaker 3 (33:13):
Institutional problem of folks just being burnt out and not
having a lot of resources for themselves that is taking
us away from the labor of establishing our nourishing human connections.
Speaker 2 (33:23):
And that's like the point of being a human. So
I think that we were.
Speaker 3 (33:25):
Just out of whack of why we're all here in
the first place, if that makes sense. I don't know
if that answers your question.
Speaker 1 (33:31):
No, it does. And I think there's this kind of sociological, anthropological,
I mean, the technology is the way into this story, right?
And it's a tool that people use. But the issues
and problems that people are facing that you're surfacing in
this book and diagnosing are not because of the technology.
The technology kind of reflects the lived experience in some ways,
(33:52):
or the way people use the technology reflects their experience.
Speaker 2 (33:55):
Exactly that.
Speaker 3 (33:56):
And I think when we get hung up on, is
this technology good or bad? We miss that more important,
more complex question as well.
Speaker 1 (34:05):
The book is Love at First Prompt, AI and the
Future of Intimacy. And also do check out Bridget's podcast.
There are no girls on the internet.
Speaker 2 (34:13):
Bridget, thank you.
Speaker 3 (34:14):
Thank you so much for having me. I'm such a
big fan of Kaleidoscope. Y'all have entertained me and made
me think for so long. So thank you for all
the amazing work that you're doing.
Speaker 1 (34:23):
I'd love to hear it, Bridget.
Speaker 2 (34:25):
Thank you.
Speaker 1 (34:25):
And hopefully you'll come back on soon.
Speaker 2 (34:27):
Anytime.
Speaker 1 (34:45):
For Tech Stuff, I'm Oz Veloshin. This episode was produced
by Eliza Dennis and Tyler Hill. It was executive produced
by me and Julian Nutter for Kaleidoscope and Katrina Norvell
for iHeart Podcasts. Jack Inslee mixed this episode and Kyle
Murdoch wrote our theme song. A special thank you to
all our listeners. Please rate, review and reach out to
(35:06):
us at techstuffpodcast.gmail.com. We love hearing from you.
TechStuff News
Hosts And Creators
Oz Woloshyn
Karah Preiss
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.