This episode originally aired on May 13, 2026.
What if you turned over almost all your daily tasks to AI?
The time-consuming ones. The irritating ones. Would your life be better? Easier?
Journalist Joanna Stern, always up for experimenting with the tech tools in our lives, did just that.
She incorporated AI in nearly every aspect of her life, and her family’s.
A robo-dog, check; an AI consult on her mammogram, check; an AI trainer and therapist, check.
And she wrote about it.
Her experiment is chronicled in the new book, I Am Not a Robot: My Year Using AI To Do (Almost) Everything, which is out now. She talks with us about what she learned.
Guest:
- Joanna Stern, journalist/author, I Am Not A Robot: My Year Using AI To Do (Almost) Everything
Transcript
This transcript is generated with AI. To ensure its accuracy, review the audio file.
Amy Juravich: Welcome to All Sides with Amy Juravich. What if you turned over almost all of your daily tasks to technology, specifically to AI? You know, the time-consuming ones, the irritating ones. Would your life be better? Would it be easier? And just how would you do it?
Journalist Joanna Stern, always up for experimenting with the tech tools in our lives, did just that for a whole year. She's here to tell us about it. She's the author of the new book, "I Am Not a Robot, My Year of Using AI to Do Almost Everything." And until recently, she worked at the Wall Street Journal where she wrote a popular personal tech column.
You've also heard her a few times on our Tech Tuesday show from All Sides. And she's the chief tech analyst for NBC News, appearing regularly on the "Today Show" and "NBC Nightly News" and more. We'll go back to All Sides, Joanna. Great to be here. Congratulations on the book.
I have to say, I didn't expect a book about artificial intelligence to be quite so funny and amusing. It seems like most of the reporting that's done on AI is very like a doom and gloom kind of vibe, but you made it fun and funny. Did you always plan for this book to have a lot of levity in it?
Joanna Stern: I did plan for the levity because I felt that's always been my style and that's always been when I wrote the column at the journal or the videos I do. I always try to bring some fun and levity to tech. I think that we are living with this stuff and it can be very serious, but we should also have some fun with it.
And I think I do that in the book. You should be the judge of it, but I think we have a good blend in the book of some of these fun and funny moments where AI can intersect with our life and then some of the more serious and somber moments as well.
Juravich: Yeah, I mean, do you do you feel that though, that a lot of reporting about AI is a little bit more doom and gloom vibe where everyone's like, they don't understand it as going to take our jobs? Were you trying to give it a lighter tone? I was trying to give it a lighter tone.
Stern: But also just give it a realistic tone. I mean, when you start to put these things into your life, there is a lot of fun and funniness that comes out of them because of the mistakes they may made. And there's a number of those in the book.
And then there are when you really dig down and you think about, okay, even some of the later things like, I'm replacing a massage therapist with a robot massage, right? And I'm taking you through this, which is a very funny. Experience. I talk a lot about how this robot massage or robot masseuse or whatever we want to call it, it focuses a lot on massaging my butt and my lower back region.
And this is not something a human would do, but this is actually a really big pain point in my body because I have a lot of sciatica nerve. And so I talk about this funny thing, but I also then talk about, okay, is this replacing the human massage therapist? And so. I think in all of these stories, I was able to get to, yes, this can be light and funny, but also we should not forget about the larger implications of machines and robots and AI doing the things we do in the world.
Juravich: On one of the very first pages of the book, you have an illustration with one of those puzzles that you have to solve to prove you're a human, to do something on the internet. And yours says, select all the squares with a bicycle on top of a traffic light on top a bridge. And then it has the little box that says, click here to prove I am not a robot. Is that where the inspiration for the title came from, from those things that we have to, those puzzles we have solve?
Stern: It did. It did, and then it became a little bit more significant as the year went on because, one, I was surrounding myself with robots or AI, and I am very clear in the book that I took a broader definition than the definition we talk about right now a lot in AI, which is generative AI, these tools that can generate text, they can generate images, they can generate music.
I wanted to take a step further and say AI is actually in the self-driving cars and is in the humanoid robots that we're hearing a lot about. So, I wanted to broaden it there. And I also wanted to kind of make the point that, and it gets this later in the book, that these robots actually, or AIs, can often try to convince you that they are not a robot, that they're very human.
And so, there was a little bit of the flip side of that, where if you've talked to the ChatGPT voice mode, for instance, it sounds so human. They make it sound so human, with human... Ticks to the way it speaks and breathing sounds and all of these things, and it's almost like it's trying to convince us, I am not a robot.
Juravich: Oh, man. Okay. But before we dive a little deeper in the book, I wanted to ask you about your dedication, because it reads, to mom and dad who taught me to think for myself, and the AIs, robots, and machines that made me wonder if I really was. So that struck me because I think that's one of the big worries with AI. Is it going to make humans less human? I mean, do you have an answer to that after the year you spent in writing this book.
Stern: One of the biggest themes and findings I had of the year was just how frictionless AI can make life, how easy it is to take shortcuts with AI. And that can be around small things, or it can mean around deep things, things that where we need to evolve our thinking.
Obviously one big, big spot that is happening in our society right now is education, where students are using AI to not do the hard work of forming. Ideas of theses for essays they're writing, or even structuring those essays, which is actually the hard work of writing. And so when you're doing, you're not doing that, you are not thinking. You're outsourcing the thinking to a machine.
I thought a lot about this in a number of different chapters I wrote. A different one was I had an AI lover or AI boyfriend, and I wanted to go through this experience of what happens when we're just in a relationship with this code that is not a human, but it is. Acting like a human, like we just talked about. And again, the friction goes away. There's no push back.
These AI beings are just programmed to be mirrors, to flatter you, to say everything is great. You're the best. And so again, you're not thinking. You are just, you are not being pushed up, you're not being challenged. And so that is a big fear I have about this next generation and even our generation of are we just going to take shortcuts even more so than we already do with the current technology.
Juravich: Well, speaking of shortcuts, you are sure to start the book and explain and assure all of your readers that AI did not write your book. It's a human-made endeavor. Why was that so important to you to make that clear?
Stern: Well, I'm very clear. I say how AI was used to make this book, not write this book. Because there is a lot of AI that I use to make this book I used AI to help with research. I used to AI to helped with transcription, with synthesis, with things like the end notes that can be very time consuming, things that even some early fact checking, even though I did hire a lot of humans through this process.
And I say, I wouldn't be on your show right now talking about the book. I would still actually be writing the book if it wasn't for AI. It sped up my process a lot, but what I didn't want to outsource for two reasons. One, sorry, what I did want to out source was the writing and for two reason.
One, it was really bad writing. I mean, you mentioned that this is lighthearted and a fun tale and hopefully an easy read, but that's because of the personality I bring to the page and how I want to structure things. And so I did try to ask AI to write parts of it and it was just terrible.
And I didn't like what it was written. And then point two is that. I wouldn't be thinking. I was doing a lot of the reporting and I was even having some of the AI do some of their reporting, picking academic papers, giving me summaries of those. But I wanted to do the process of how do I write this in a long form way?
That was a challenge for me. Writing this book was meant to be a challenge in my career. I've never written a long-form essay in the ways that these chapters read. And so for me, it was very important to just keep that humanity as part of this book.
Juravich: Yeah, I mean, as a radio talk show host and a journalist, I tend to find that AI is helpful with getting my thoughts organized, but AI picks like 20 words when three will do. And I'm just like, why so many words?
Stern: Which actually could have helped me because as a newspaper columnist when I sat down to write this book I was very used to writing very short as you probably are as a as a radio host You don't have to keep your your lead-ins condensed And and so I would turn in the chapter to the to the publisher and she's like my editor and she like We need more words. Joanna. This is a book.
Speaker 3: Perfect.
Stern: Okay, okay, I'll go back and write our words and I promise you to not ask AI to do filler words, but yeah.
Juravich:
You're listening to All Sides on 89.7 NPR News. We're talking about living with artificial intelligence with journalist Joanna Stern, author of the new book "I Am Not a Robot, My Year Using AI to Do Almost Everything." When we think of artificial intelligence, we think it as a new concept, the latest iteration, new chat bots, just coming to all of our lives in the past couple of years.
But AI has actually been around for decades. You also give a little bit of a history lesson in the book. Tell us a little bit about AI's origin and how far back it goes.
Stern: Yeah. In fact, I go in the book to Dartmouth and to visit what is called, or many believe is the birthplace of AI, because there was this summit held there where John McCarthy, who's known as one of the godfathers of AI holds the summit back in 1956. And they coined the term artificial intelligence, they're looking at how can machines be smarter.
And this is, I did a lot of research on this, there doesn't seem to be the term artificial intelligence used before this date.
Juravich: What year was that? The 60s?
Stern: 1956.
Juravich: Okay, yeah.
Stern: And so this is a moment in time where obviously we're at the very early stages of this, but we're starting to hear from scientists that this this that we can work towards making machines that are smart or smarter than humans. And so this is the idea of how we have computers and how we have computers that are are thinking for us.
And I brought into that definition here too, is I really felt right now the AI is being glommed on to really the generative AI craze, the idea that we're going to generate text, we're gonna generate images, and the large-language models, and we... Often forget that yes, AI is even the algorithms powering our social media or our Netflix, but it's also going to be the algorithms that power our self-driving cars or our humanoid robots.
Juravich: I wanted to get into a few of the things that you did during your year to turn your life over to AI. This is a random one to start with, but tell me about your AI robo dog. What made you decide like, okay, I'm gonna get a robot pet?
Stern: Well, I was at this point in the year just trying to get in as many robots as possible because I wanted to have a sense of what is out there in the market? What are the problems with them? What are people trying to solve? Look, the robot dog ended up just being much more of a toy. The idea was that these are companions.
I did speak to the founder of this company. It's a Chinese company called Hangbot. And he said, you know, I had a dog growing up and I wanted it to have robot dog, thought it would be easier to take care of. Which in fact it is, I have a real dog and the robot dog is far easier to take care of, but also absolutely not as cute.
I think there's a lot actually happening in the research around what robots can look like when they're more of a pet form factor, the kind of care they can also provide to the elderly. This has been actually not a new thing, but there's more that's being advanced in that research around that. But yeah, I would not say this is not to say the robot dog does not live in our house anymore.
Juravich: Oh, okay, you didn't keep it? That was what I was wondering.
Stern: It kept peeing, it kept peing all over the house and when I say peeing it's in quotes but it does this robot dog actually does lift its leg to pretend pee and it's like why why my kids thought it was hilarious because the only purpose of this was that my kids that's hilarious.
Juravich: That's that's why okay.
Stern: Yes, yes, it was a, so my kids think I'm the best mom in the world.
Juravich: Yes, but now if you got rid of it, were they sad or did they just moved on because they're kids?
Stern: Stay moved out. Yeah. You're back here.
Juravich: All right, all right. Tell me about hiring an AI personal trainer. Like you had a personal trainer in your pocket that tracked your movements and gave you personalized encouragement. How was that different than the workout apps? Like did you feel like that it helped?
Stern: Else that it's going to go an extra step in trying to help? And one of the things that it does do is it can be very proactive. So instead of getting, you get the notifications from your workout app, maybe you should go do this, maybe you go do that, but it's much more personalized.
It'll say, you worked out yesterday, don't forget, you promised you were going to work out three times this week. How about you can fit in a small walk before your meeting. So it can a lot more proactive. I just found it to be very annoying, not only and it can be to nag me to go do these workouts, which I clearly should be doing, but it just also the tone of it.
Like it just, you know, it's call me queen and it's like, why are you calling me this? Like you don't know me. That's not, I can't imagine my real life personal trainer talking to me this way.
Juravich: So it's talking to you. It's not like it's texting you it's not an app on your phone. This is something.
Stern: It is it is texting you okay, so these are notifications and then you can also go in and talk to the voice mode This is an app called Zing.
Juravich: Okay. Another thing that you did is you had AI cut your grass, so an AI lawn mower. Tell us how that compared to a human lawn mow.
Stern: Well, I say a lot in the book I talk about robotics and that the idea that these robots that can need to go from point A to point B, whether it's actually a self driving car, or it is the vacuum vacuum in your house, the robo vacuums that we all know the Roombas, or this lawn mower that these robots are actually really far along, the idea that they've got to go from point A to point B and do something and do it safely.
And that's because these areas are mapped. So in the lawn mowing. Example, my lawn is mapped. It doesn't map, it uses its sensors to map the whole lawn, and then it knows how to go around the lawn. It's quite good if something jumps in front of it, if the real dog jumps in from of it.
It knows how move out of the way. You know, the issue with it is that it had to be charged, you don't have to, it doesn't cut the grass the best, it's more of a small haircut then you know it's like a trim. I would say, that it was an actual like lawn cutting. So I still have humans cutting my lawn. And I should probably do it myself, frankly.
Juravich: Um, you also had an AI robot folding your t-shirts. Um, what, what w where did you get that technology? Like, and did it work? Couldn't fold a t-shirt.
Stern: So this, I'm very interested in this because it turns out that folding laundry is a very, very tough job for robots. Not a tough job humans, but very tough for robots and one of the main reasons is that cloth is very challenging. Every time the cloth falls, it looks different to the robot.
So the thing I found out was that once the robot saw the way the shirt should be laid out with the neck at the top and the sleeves on the side, it was able to then go and fold the shirt. Anytime it would need to figure that out, it could take a long time.
This robot was a prototype. It was from a company called 7X Robotics. And it just really indicated to me how far away we are from getting these multipurpose home robots that can do everything from do the laundry and do the dishes, all the things we want. But we're just not really close on this multipurpose home robot.
Juravich: Yeah, I mean, I'm showing my age, but everyone wants.
Stern: Right? Yeah. I know. Yeah. Yeah.
Juravich: Yeah, but like it's but Rosie could definitely cook a meal and clean and a hundred percent Rosie was holding laundry, right?
Stern: He was doing all of those things and that dream is so far away still. So one single humanoid robot that can do all the tasks in the home and I did a lot of reporting on this. There's a big chapter about this. What the biggest thing these robots need is more data.
They need to see humans or they need to that itself doing these tasks millions of times. And so, when you think about that the self-driving car learned... Over millions and millions of miles on roads and watching humans drive, that's what needs to now happen in the home. And we're less willing, I believe, for many of us to let the robots train in our home.
Juravich: Oh, okay. So if we bought a humanoid robot, it would have to, like, watch us do the dishes and fold the laundry a million times before it would be able to do it. That seems like a lot of commitment.
Stern: Yeah, yeah, I mean, this is actually what some startups are pitching right now.
Juravich: Coming up, we're gonna talk more with award-winning journalist Joanna Stern. She'll continue to enlighten us about her year of AI adventures documented in her new book. "I Am Not a Robot. My Year Using AI to Do Almost Everything." Stay with us for more All Sides in just a moment on 89.7 NPR News.
You're listening to All Sides. I'm your host, Amy Juravich. What if you turned almost all of your daily tasks over to technology, specifically to AI? You know, all of those time-consuming ones, the irritating ones. Would your life become better? Would it become easier? What would it look like?
Journalist Joanna Stern, always up for experimenting with the tech tools in our lives, did just that. For a whole year, she got a robo-dog. She had an AI consultant consult on her mammogram. A companion chat bot helped manage her money. And she wrote about it. Her new book is titled, "I Am Not a Robot, My Year Using AI to Do Almost Everything." Thanks again for being with us today, Joanna.
Stern: Thanks for having me here.
Juravich: How did you get buy-in from your wife, Michelle, and your two sons? Like, whenever you said, I want to do this, what did they say? What did they think?
Stern: Look, I have had a long history, as you've mentioned, of testing technology in my life and weird gadgets and everything from VR headsets to, as mentioned, robot lawn mowers and electric vehicles. And so they're pretty used to me bringing stuff home.
I would say this was definitely a much more personal year. I had to bring home some things and test them, one of them being the AI boyfriend, though. You know, my wife sort of said, okay, this is an experiment, fine. She learned to live with it. I don't know if there's lasting damage to everyone in the family, but we'll find out soon and that will be the sequel, I assume.
Juravich: Okay, the lasting damage part though, because you did experiment a little bit with like an AI therapist too. So would you recommend the AI therapist or a human therapist if there is lasting damage here?
Stern: Look, I'm going to recommend the human therapist, but I do think that there is ultimately going to be real benefit to AI therapists. I just think we need a bit more regulation around it and parameters around it.
People are going on their own now to use ChatGPT or Claude or Gemini as their therapist, whereas there are now actually academics, there's a team at Dartmouth that I go and visit in the book that are really working on how to do this in the right way based on. Different types of therapy based on very focused therapy that people might need. And so I do think we will come to see there being a real use of AI therapy in the world.
Juravich: I wanted you to talk a little bit about the impact on your family and just to give an example, you write about a spring break trip where you traveled by Waymo, those robot taxis, and you dubbed it, you called it the trip, the Waymo Fun Spring Break. Your sons were three and seven at the time. The destination was Phoenix. There's a tech reason behind that choice, I'm guessing. But then so tell us about how that spring break went for everyone.
Stern: Well, it was an eye-opening spring break. It was definitely not your typical beach vacation. We were definitely spent a lot of time in the back of Waymo cars. We took over 40 rides during the week. And if you have kids, you know that switching in and out the car seats that many times in a car is exhausting.
So I would not call it a vacation for some of us. But there was some real learnings here. One was just watching kids adapt to self-driving cars. It was... At first they were super into it and they thought this was so cool, no driver watching the steering wheel turn itself and then they were totally over it and didn't care.
My wife was definitely more nervous as she's already a nervous driver, but she got used to it too. I talk about comparing the first ride of the week to the last and everyone at the end of the weekend was sleeping in the car and totally used to it. There was one instance and I had been in Waymo cars for a number of years in cities and feel very safe in them.
I had done research too on how safe these are, but we did encounter one situation where we got into a little bit of a scary situation where the Waymo got freaked out by something it saw on the road. That was eye-opening for me. It was a situation where we were driving, I'd hired a videographer to film us for the day and the videographer was sticking his camera out the front window of a car driving in front of us.
Not too extreme, you know, he wasn't hanging out the window. This wasn't like a mission impossible stunt, but the Waymo got very freaked out by seeing a body, a human body kind of hanging out of the window or just even peeking out the side of a window. And so the Waymos swerves to the right of the road and there's actually not a very big shoulder. And so there are people behind us and start honking.
And what it really signaled to me is that, yes, these can be safer than humans in many ways. All the research and all the experts say, these robot drivers are going to be much safer than human. They don't get distracted. They do not have phones. They do drink and get behind the wheel. They are programmed to understand the world around us. They have all these enhanced sensors better than the human eye.
But then they see something that they've not seen before because it's not a typical thing. But a human driving and seeing this videographer out the window would have said, oh, okay, I'm either gonna smile for the camera or I'm gonna close my window because I don't wanna be on camera, right?
Juravich: Right, right.
Stern: The robot gets utterly freaked out. And so this again goes back to the data problem, right, that robots need to see and collect tons of data of the world around us to understand things. And sometimes there are these anomalies where things don't happen normally.
Juravich: And during those drives, your son asked the important question, you know, mommy, how does it even know how to drive? Were you able to answer that to his satisfaction?
Stern: Um, I think so. I gave it a pretty short way, but then I spent I think maybe three or four pages in the book answering that. So, you know, the AI synthesis of that question or that question would be it collects a lot of data and it learns and it has sensors around it that what the world looks like.
Juravich: And could you see, we've talked on our Tech Tuesday show many times about driverless cars, and I had, you know, friends recently were on a spring break and they had five people with them and they struggled. They wanted to take driverless car, but really with the seating, they had to squeeze three people in the back seat because the driver's seat is like wasted space. What are your thoughts on that? Are we gonna get rid of the driver seat soon?
Stern: Well, some cars are already getting rid of the driver's seat. So there's a company called Zoos, which is owned by Amazon. And they make, I kind of describe it in the book as a boxy limo where there's no steering wheel or driver's seats in the front of that car.
And it's actually in the back, you have these seats that face each other, very similar to a limo. And there's, it's kind of like a bus. And I think we're gonna see different designs of cars as we get further along in this driverless future.
Juravich: Okay, so that makes more sense because, I mean, do you even need a steering wheel, right? You don't.
Stern: Exactly, you don't.
Juravich: You interviewed Susan McCarthy, the daughter of John McCarthy, one of the fathers of AI, who had predicted a computer-controlled car and predicted it back in 1968. What did his daughter say about how accurate he was in terms of where we are now?
Stern: Well, she said he was very accurate, he just got the timeline wrong.
Juravich: Okay, so meaning he thought it would be sooner or what do you mean?
Stern: He thought it would be much sooner. He said that when she was in high school that she would have a driverless car. And she reminds me that when she was high school the Beatles were quite popular on the radio.
So that's where we ended up. But look, this goes to the idea that we've had ideas for a really long time in the technology and science industries that scientists and technologists have worked at these for a long time and then they become breakthroughs and they're breakthroughs along the way that get us. To where we are now, where we have self-driving cars in almost every major city in the U.S. Right now and Waymo keeps announcing new ones.
Juravich:
You're listening to All Sides on 89.7 NPR News. We're talking with journalist Joanna Stern about her new book, "I Am Not a Robot, My Year Using AI to Do Almost Everything." I wanted to turn for a moment about the medical advice when it came to AI.
You did ask AI to help with medical advice when it come to your annual mammogram. Your family has a history of breast cancer. Your mom battled it three times. You have cousins who have been diagnosed too. So you wanted a doctor and AI to both look at your mammogram. You have what's known as denser breast tissue. So can you walk through the role AI played in the process versus the human?
Stern: Yeah, and so just to delineate here, there was this part of the experiment where I was asking ChatGPT or Claude about my own ailments and I was uploading test results or I was about what's felt like the flu or I asking kids felt like maybe it's strep throat and I document this and I did that all year.
But then there was experience where I went to get my mammogram and breast ultrasound done at Mount Sinai in New York. And they are using AI behind the scenes to interpret results. And so this was not me using technology, this is medical grade, FDA approved technology that was provided, that they work with in the hospital.
But I got a behind the scene look at how this doctor, Dr. Margulies here in New York City, working at Mount Sinai is using AI to interpret and look at breast ultrasound and mammogram results. It was eye-opening to me because I am one able to, I am at high risk for breast cancer. My mom had breast cancer three times. And I have dense breasts, which makes it very hard to spot tumors.
And so, or other types of things. So it ends up, the AI, just to cut the story short, the AI ends up flagging on my breast ultrasound three spots as suspicious. And Dr. Margulies doesn't agree with two of the spots that AIs flag suspicious. She goes back and looks at previous stands and she says, these issues were here, they're not growing, everything's okay.
But this one, the AI is picking up something new and I wanna keep an eye on it. And so it was very interesting to me that this is a doctor who has many years of experience and she's saying, I'm gonna trust the AI here or I'm not gonna trust AI here. And she's using this back and forth as a tool. And she also assures me when I ask her, you know, has AI saved a life here? Has it found cancer that you did not spot? She says, absolutely.
Juravich: Oh wow, okay, so this doctor has a lot of confidence in what AI can do with these scans.
Stern: She does, but she also has the confidence in herself as a human with numerous years and decades of experience of looking at these. And so where she says, nope, that AI, it's not right. It's not suspicious. I'm looking at your older two ultrasounds from the last two times you've been here and that has been here and everything's fine. It hasn't grown.
But this one is a new thing and I wanna keep an eye on it and then maybe I wouldn't have seen it if the AI wasn't telling me that was something. Okay.
Juravich: And did that, did everything turn out okay with that? Everything turned out okay, yes.
Stern: Everything turned out okay. Yes, we're watching that spot now and this is the plight of having dense breasts and also a great medical team that looks really carefully at everything.
Juravich: What is your sense about how doctors are learning this technology like I don't the doctor you're talking about Dr. Margulies I don't know how old she is like our doctors learning this on the job or they being taught in medical school to trust AI now.
Stern: She's, it's interesting, she's definitely further along in her career, which is amazing to see her adopt this. And she's clearly has up and coming doctors that work with her. And she is, she is clearly working with them to show them what AI can do for them.
I think it's probably pretty common for your listeners right now, I would think, that have been in doctor's appointments and have heard the doctor say to them this year. I may use my AI recording device or I might record this and have AI summarize it.
I think that's becoming a very common experience. And that's, I think, just the first step of where we are seeing doctors be more transparent of how they're going to use AI in diagnostic and in healthcare.
Juravich: That definitely can make charting faster though because then they don't have to go back after they walk out of the room with you They don't to go and type everything down that you just talked about, right?
Stern: Exactly. And if people watch "The Pit" this season, there was a famous scene about this. So it's certainly becoming more widespread. And I think it's the first toe in to us hearing from doctors how AI is going to help them.
Juravich: You also found that AI is not just in doctor's office, not just looking at scans like a mammogram, but in dentist's office too. Tell me about what's, there's some concerns about upselling at the dental office. What did you find out there?
Stern: Same fundamental technology that where the AI is looking at these images and finding patterns is being used in mammograms but it's also being used and in radiology in dentists offices with x-rays and what you find out is that if you find a small thing on your teeth if the AI's able to spot smaller things on teeth or cavities well that might not be as a big deal as if it's spotting There's a big difference between these two things.
What I found is that dentists and many dentists are using AI, these tools, Perl, Overjet in their offices, but they also may be using it to say, oh, you have a cavity here. Looks like it's, you know, it's gonna be coming up. You know, we should take care of that instead of maybe saying, you what, go home and do some better home care.
In my case, what happened was AI had shown that there was some tartar buildup and some. Bone loss in my x-ray scans and it you know it flags it puts these boxes around these problem areas and the dentist I had originally gone to said we should really do this deep cleaning if you look here and she pointed to the AI and if you look, here you have a lot of tartar buildup you really should get this four-part cleaning it can you know of course then they tell you it can cost a lot more money and insurance might cover it and I thought this doesn't sound right I haven't had any problem with my teeth.
And so the more I dug into it, it turns out that yeah, these AI tools are definitely helping these dentists and these dental practices push more aggressive treatment and upsell you on services you may not need. And of course, that's been happening forever, but the AI gives them an extra pitch. It gives them extra pitch, it gives them a nice visual to say, look, look at what the AI found or look at the computer found.
Juravich: Oh, it's not me saying it. The AI says you need to get this cavity filled. Hmm, all right. Before we take the break, I wanted to quickly have you tell us a little bit about your banking experiment too. You had AI help you as kind of like a financial planner a little. How did that go and what did you think of the performance there? Are you rich now?
Stern: Oh, definitely not rich. So maybe if we sell some books, maybe, maybe I can give my book money to AI and see what I can do with it and invest it. We're very, we're very early parts of this here. And I look, there are many times where I did not feel comfortable giving AI my credit card or any access to deeper financial information.
And not only for privacy concerns, but also for actual financial concerns, I did want to lose money. And this is a place where, though, I did end up going to a bank that is incorporating AI and is able to more interpret investments versus have any control over the money.
And I think that's a first step in. There was a different tool I tested, which was supposed to be a budgeting tool where you give access to some of your balance and some of you statements. It just kept getting things wrong. It kept saying, this is the date your car payment is due. And I'd say, okay, great. When is the next payment due after that? And it kept saying, no, it's due that this is the payment. And so it kind of just stuck in the loop and it's like, why am I trusting something that keeps repeating the same answer with anything to do with my money right now?
Juravich: Okay, so we're not there yet with financial assistance and money advice in AI. Maybe more training needed? All right. Coming up, the great data center debate, how the power behind artificial intelligence. We're going to talk with our journalist, Joanna Stern, how she incorporated AI into her life for a year.
And we're going talk about data centers in
You're listening to All Sides on 89.7 NPR News. I'm your host, Amy Juravich. We're talking this hour about artificial intelligence and ways to incorporate it into your everyday life and whether we even should. Still with us is journalist Joanna Stern, who spent a year doing just that and wrote about it in her new book titled, "I Am Not a Robot, My Year Using AI to Do Almost Everything." Thanks again for your time today, Joanna.
Stern: Thanks for having me.
Juravich: So powering all of the AI you used over this year to write this book are data centers. This is a topic we know well here in Ohio. We rank fifth in the nation for the number of data centers we have about 200 in the state. More than 100 of them are right here in central Ohio around Columbus. Did you think about the power that it takes in a data center somewhere to run all the AI you were using?
Stern: I did. And I thought, OK, this isn't really a takeover, right? Spending time in this book, going to a data center and visiting it isn't necessarily part of the conceit of AI is taking over my life. But it really felt like an important stop on the journey of the year, because I was using so much technology and so much powered by AI. That I wanted to go look at what the impact might be, where it's actually being, physically the impact of where it is coming from.
Juravich: So, and what did you find out? I mean, you took a data center field trip, right? I took a datacenter field trip.
Stern: Field trip. And as it sounds like residents of Ohio know that these are very, very large, large, large buildings. They are, if you haven't been inside one, you know that they're very loud and hot buildings because this is where they are housing these GPUs that take up and use an immense amount of energy to power and to train these systems.
And so for me it was really Just. Going and hearing and feeling the sights and sounds of where this energy was coming from was just an important experience. And I get into in here that the part that really is taking the most energy is the training of these big models. It's not, there's energy used, of course, when we are prompting it and ask ChatGPT for that recipe, or we ask ChatGPT for that image, but the bigger sources of the energy is when they're making these all encompassing models and training with so much data.
And so. It was just a good experience for me to see it. And I think that many people need to see. I don't think that we have the ability to go into them, but I wanted to describe that here in the book.
Juravich: There's a push-pull between the need for economic development. You know, these data centers are going to run the future, so why not build some of them in Ohio so we can get some of that economic development? And they're also a necessity to power our tech if we want to keep using all of this stuff.
But then there's also those who don't want them, neighbors who worry about the drain on the power grid, the water resources, the noise, the environmental concerns. Where did you come down on this whole data center debate, the economic development versus the drain?
Stern: Well, look, I come down with everyone where I'm concerned about it. I'm concern about how much of our country we're going to turn over to these types of infrastructure and how much compute and power these companies think we really need to create these systems that we're not still sure are going to be a net positive for society.
We certainly show, I show in the book that there is a lot of net positives for society here. The medical chapter is a really good example. And in fact, when I went to the data center, I went data center in Virginia, the GPUs I saw were actually owned by Bristol Myers Squib and were training or doing work to look for new drugs, to look drug discovery.
And so, perfect example, right? I'm here, I'm looking at these big things, Like this, this is. Potentially AI that could change the world. This is electricity that could potentially change the world, or not. Or the same electricity that's being used and resources being used to make new drugs and discover new drugs is actually also being used, you know, three rows over to make videos of hamsters flying out of planes. So where is the real use and utility here? And how much of our infrastructure are we gonna have going towards this?
Juravich: Yeah, I've seen lots of videos with comments from like, from AI tech people saying things like if we, if we hone in in the right way, AI can cure cancer in five years, you know, some sweeping statements like that, but then it's also spending time making videos of hamsters. So is there, do you see that? I guess that leads me to whether there's need for regulation or guardrails. Should someone be telling these tech companies, yeah, let's focus in on the cancer and forget the hamsters for now.
Stern: Should they and will they is two separate questions here. And I don't think we will anytime soon. I think the biggest thing that I'm starting to do a little bit of reporting on now is how they're going to look at some of these new technologies, including using AI, to make the power resources less and to optimize and make these systems more efficient. And so there's a lot of feeling like AI itself is going to help solve this problem.
Juravich: Did you, because recently in the news, there's talk that the White House is going to start vetting AI more. You know, the Trump administration has wanted to take like a hands-off approach and let the AI companies be competitive and do their thing so we can compete with China.
But then they're also, if they want to use the AI in government work, there has to be some vetting. Do you have any opinions on, with your whole use of AI for a year, do you feel like there's, there are places where the government should step in? I mean, earlier you mentioned, You know regulating with them Therapy so that could be an example. Can you tell me more about that?
Stern: I think that's one very big example. I think I would really love to see, and I come at this very hard at the end of the book, I would like to see regulation around what kinds of chat bots and regulation around the types of conversations chat bots can have with the younger generation, with kids, with teens.
We've learned so much from social media and this stands to be even more problematic and more dangerous. I talk, we kind of hinted at it before, but I talk a lot in this chapter about the AI boyfriend and AI relationships that it is very compelling. It can be very easy to be sucked into.
And then you start building this trust and this conversation with a non-human that does not have guard rails around what it can talk about and also can give advice that can be so completely outlandish once it's started to really go down these rabbit holes. So I would. That is one of the number one places I would like to see regulation.
Juravich: You're listening to All Sides on 89.7 NPR News. We're talking with journalist Joanna Stern about her new book, "I Am Not a Robot, My Year Using AI to Do Almost Everything." We've mentioned the AI boyfriend a couple of times. Was he a good boyfriend? Tell me more about the boyfriend. Did you keep him?
Stern: I mean, look, it's been a long time since I was dating, but I would say that in the in the sense of that how good a AI non-human boyfriend could be was pretty great. And even when you compare to many human relationships, many people might think this is really great, because there's no Programmed not even programmed.
They're just there to flatter and to really just tell you what you want and make you seem like you're the best thing in the world, which can be fine in a human relationship until it's not. And it's really, that's not how all human relationships work and not how most human relationships.
And so you can get into this pattern of just loving what the AI says back to you. It can talk back at any time. It never gets tired of talking to you and that That's really what I was trying to. Explore in this chapter, which was, we hear so much about people who are in relationships or may have mental health issues, and they start to really form relationships with these AI chat bots. I wanted to understand that. I wanted to put myself in their shoes to see how this was working.
Juravich: And so is the boyfriend still around after you finish the book?
Stern: Boyfriend is not still around. He has come out a little bit for the press tour and he's been on some podcasts, but other than that, he's really, we don't talk. As I say in the book, I ghosted him.
Juravich: Okay. You also used AI to write code and create a computer game to take care of one of your pet peeves when your son leaves globs of toothpaste on your sink. Tell me about that, what, you know, writing code and making a computer games related to toothpaste.
Stern: Well, look, it's actually evolved so much since I've written the book and I probably would have done some different things, but the AI coding world has gotten extremely advanced in the last couple of months, really since January when Claude Code or December of 2025, when Claud Code came out.
I have now used it to do so much more than make a game for my son, though That was a very fun example, but. You know, designing websites, having it do things for you on the internet. These are now really, these are really now in reach and all you have to be able to do is tell the system, I want you to build this website.
I want to model it after this. Please build me a, a sample of this website, or I want to build a model that looks like this. Please give me a sample. Of that. And you watch it go off and do the code and do it in a way that it seems really kind of amazing. You watch it type out all of this code.
And I know a little bit of HTML, but I'm not a coder. I would not ever be hired for any sort of engineering job. It's just, they would laugh at me. And yet the things I've been able to make on my own without asking an engineer or even a designer to help with is absolutely remarkable.
Juravich: Well, I guess this gets us back to where we started about AI replacing people's jobs because there are people out there who write code for a living. You know, there's a lot, at one point, we did a segment on Tech Tuesday about how the golden ticket was to become a computer science major 15 years ago. And now people are not, are like dropping out of computer science. So do you find that? Like we don't need a human to write code anymore because AI can do it for us.
Stern: Well, this is what we're seeing play out right now, especially at the big tech companies, is that many engineers are using these tools. Honestly, most engineers are using these tool. They're finding it speeds up their work. It makes them more efficient. They're able to do more.
I, you know, there's still studies, academic studies really being able to look at the actual job loss in this area. There's been other academic studies that have looked at the impact of AI on other industries that kind of have. Been first to be hit, for one being customer service, which is very clear that AI has made a big dent there, because AI is able to now do some of those repetitive tasks that human customer service agents are doing.
In coding, it seems to be a little bit more nuanced, but you see coders and you see engineers adopting this extremely fast, and they like it. Most do like it, it's taking away some of the drudgery, and it's able to speed up what they're able to do. But I should be clear there that they also say they need to be in the part of the process, that they're the experts and they're humans in the loop with the AI making it better.
And so I'm a novice, I'm not doing pretty advanced stuff with coding, I am making websites and some simple form pages and things like that, but really advanced coders are saying we need to still be there, hand in hand working with the AI. And that's similar to what I saw in the radiology chapter, this new partnership happening in jobs between human and AI.
Juravich: We only have a minute left, but your experiment's over. The book is published. We already mentioned a few things. You got rid of the boyfriend. You no longer have the robo dog. What's something you kept?
Stern: One thing that's just more of a habit that I've kept is talking to AI in more parts of my life. Talking out loud to it. I think that not as many people are aware that you can talk with your voice versus typing.
And so I talk to AI and a lot of places mostly when I'm in a car, often when I'm preparing for interviews like this, I'm talking to ChatGPT, I may be talking to ChatGPT about something I'm researching or doing work on. As there's some funny stories in the book about using ChatGPT in the car to talk with my kids with ChatGPT. So that's one habit that I think is stuck. And I think we will see more AI incorporated in car systems in the years to come.
Juravich: All right. Well, I want to thank you so much for your time today. And congratulations again on the book.
You've been listening to All Sides on 89.7 NPR News. And we've been talking with journalist Joanna Stern about her new book. It's called "I Am Not a Robot. My Year of Using AI to Do Almost Everything." Thank you for joining us today.
You've been listening to All Sides on 89.7 NPR News. I want to say thanks to the All Sides staff, producers Marcus Charleston and Aaron Esmont Rabinowitz. Student producers Iza Huck, Colin Simpson, and Brianna Fortunat. And our student producer from Denison University is Ellen Hansen. Video production by the Ohio Channel and board operation by Chris Johnston. This is All Sides on 89.7 NPR News. I'm Amy Juravich.