Author Talks: Testing AI’s limits in a one-year experiment

In this edition of Author Talks, Christine Y. Chen of McKinsey’s Institute for Economic Mobility chats with Joanna Stern, founder of The New Things, a consumer technology media company, and Emmy Award-winning technology journalist, about I Am Not a Robot: My Year Using AI to do (Almost) Everything (Harper/HarperCollins Publishers, May 2026). Stern shares insights from a year spent testing emerging AI technologies across work, family, and daily life. She gives readers a close-up look at the challenges of full immersion—offering a road map for using AI effectively and a cautionary lens on where humans must draw the line. An edited version of the conversation follows.

How did you decide to immerse yourself in AI for a full year and write about it?

For the past two-plus years, the entire industry has been talking about AI and how it’s going to affect every part of our lives. I kept hearing that, time and time again, from AI executives, big tech companies, academics—the whole industry. At the time, I was writing a column for the Wall Street Journal and talking about AI in various parts of our lives.

From cooking to medical to various areas, we wondered, “How is technology going to really help? After many meetings with these companies, I had the idea of trying AI for a while, and then it became a year. I thought, “Why don’t I just try for a while to see how AI will affect all parts of life?”

I wondered how I could weave AI into the parts of our lives where we questioned how this technology could make things better. That was the start of the idea. Before I knew it, I had sold the book, and I had to proceed. So, in 2025 I tried to get as many AI gadgets, services, agents, and people to talk to me and to intertwine them into my life.

How did you feel as you started your year with AI?

Anxious. In the book, you read about my anxiety and about writing the book based on what’s happening in the industry. You mentioned that it was a year ago, but it feels like it was 15 years ago in AI time.

I was very nervous about documenting a year where everything is moving so quickly. I thought, “Will the problems I have at this point—or that I’ll have when the book is on the shelf—be the problems that readers have?” I’m very happy to say that a lot of it was quite evergreen.

Yes, the technology changed, but the themes didn't change. For that year, I was constantly chasing what was next. I was pushing everyone in interviews to say, “OK, imagine us a year from now.” Of course, the troubling thing is that nobody knows. Nobody knows how fast or slowly this [AI progression] is going to move. It has moved quickly across the board. So, I was very anxious about wondering how I would capture what was going to happen.

What were some of the things about AI that made you anxious?

Many things made me anxious. The thread of our personal data is carried through the entire book, and that’s one of my lessons about how we should live our lives going forward. One lesson was that I have handed over a ton of data to these companies. In some cases, there was a real benefit to doing that. In a few sections of the book, I wrote quite a bit about wearing the Bee bracelet, which captures audio. The bracelet provides a transcript of everything that happens in your life that it hears. Then it throws the audio away so you can’t go back and listen to it, but it gives you this transcript of entire days of your life.

Amazon acquired the company in late 2025. The purpose of the Bee bracelet is to make your days better, to clearly summarize the things that happen throughout your day, remind you to add things to your to-do list that you don’t remember on your own, and provide an analysis of things you can do better.

It is wild what the bracelet generates. You forget about a lot of things that happened during your day. You say you’re going to do things, but you don’t do them. It’s also really oddly reflective. I asked it to give me a glimpse of what happened to me this year, and it wrote a very poetic piece about my life, my kids, and my life with AI.

It was as if your journal could talk back to you. I received a real benefit from having this bracelet. Yet there was also a huge cost I had given the company, which is now Amazon, access to all this information about my life, and that is very much a theme as you go deep into any AI tool and any AI benefit. There’s going to be the significant cost of data. In a way, that’s widespread surveillance.

What surprised you during your year with AI?

At the Wall Street Journal and for my career, I’ve done a lot of personal stunts, journalistic stunts, brought technology deeply into my life, and tested it in very creative and crazy ways. The year with AI was akin to that, but on steroids. Any place where I wanted to integrate AI, it would have to be in my personal life as well. It delighted yet terrified me to watch AI and the future of robotics and machines through the eyes of my two kids.

I thought my children would have a fun line here or there in the book. Yet there were a couple of moments through the year, specifically during our spring break in Phoenix, where we rode in Waymo cars for the week, and I saw the future through their eyes.

I did not expect that seeing the future of machines, computers, and robots through the lens of the future generation would be as big a focus as it was. I also didn’t expect to be as moved and terrified as I was. I thought, “This car is driving!” The thought that it was a very good self-driving car doing everything virtually on its own was surprising, especially for those of us who have been in self-driving cars and a part of tests for years. 

We’re adults, and we have grown up in taxis and other means of transportation. But kids don’t really know that world. Yet the speed at which they adapted to the absence of a driver took perhaps seconds. My four-year-old fell asleep in the back seat of a self-driving car within five minutes of getting in. The fact that the car was driving itself just didn’t faze my children.

Meanwhile, my wife, who’s a pretty nervous driver, was freaking out. With somebody else in the experience who really can see you, you pick up a lot of the social or cultural cues of driving.

At that moment, it crystallized that this was going to be their life. When machines are doing things and holding their [the future generations’ lives] in their hands, it will be natural for them. They won’t think much about it because it’s already happened for them at age four.

How do you view AI today compared with other big technology transformations in the past?

The beginning of the book tries to set that parallel. For example, in the late ’80s and early ’90s, we could not have imagined someone tapping us on the shoulder and saying, “You’re going to do everything through a computer, and then eventually through a little glass slab in your pocket.”

We would have said, “That’s crazy.” The question about this moment is, will we be at an inflection point in technology where everything changes? We will tap someone on the shoulder now and say, “Your AI agent will actually run your entire business in ten years,” or “AI agents are really going to take jobs,” or “AI agents are going to change the education system.” And we say, “That can’t be.”

This book is about that life transformation and societal transformation, but through my lens and living this life, not by myself, but with my family. By seeing through those generational strides, hopefully, this next generation will be thoughtful about how we implement AI.

In one chapter, I write about returning to Union College, my alma mater. I sit in a class and quickly realize that no one has done the reading because ChatGPT summarized it.

Then I met a student named Grace who told me how much she’d been grappling with using AI. AI was doing all her research and much of her writing, but Grace was clear that we don’t want to take that risk with writing. AI is doing so much for her that she said simply, “I’m not doing the thinking.” In that pivotal moment, I realized how important the formative years are in our educational, social, and professional development—and this next generation isn’t going to get that unless they really try and unless they’re really aware.

Grace was very self-aware. She said, “I’ve stopped. I now really try to make myself think, because what are my parents paying for here at college if I’m not actually doing the work and if I’m not actually thinking?”

I have a similar reaction, grappling with thinking, “What is AI doing to me?” Yes, as humans, we do try to take the easy way out. We’re always going to take the fastest route to town, or we’re going to take the fastest route to try to get to the money or whatever it is. But we also know the inherent risks of doing that sometimes.

What was it like having robots in your life?

The book is structured in seasons. Summer was what I called “bot girl summer,” or robot summer. I wanted as many robots as possible to do things such as vacuuming and mowing the lawn. But laundry was the one task that I really wanted the bot to handle. I spend Sunday nights folding laundry; there are so many other things I could do with that time. With the promise of AI and efficiency, we’re going to get our time back. I tried so hard to get a laundry-folding robot.

There are lots of companies tackling this issue. I found a start-up that had built a pretty simple laundry-folding robot that works in homes. It is not a humanoid but has two robotic arms. The robotic claw arms hook up to a laptop that runs the laundry-folding model. It does fold, but there are issues.

At that point, it could only fold T-shirts. It folds very slowly, but it does it autonomously. I continue to report on robotics, and very few humanoid robots are now completely autonomous. We have self-driving Waymos and Zoox, and Tesla’s robotaxis, which are autonomous in large part; they just go from point A to point B with no human intervention needed. Yet that’s not the case with many humanoid robots and laundry-folding robots.

Generally, when you see a humanoid robot folding laundry, a human is operating it behind the scenes, using a controller or a VR [virtual-reality] headset to manipulate the robot’s hands. In this case of the completely autonomous robotic arms, the robot sees the laundry fall, which is the biggest challenge for the robot. I spent days thinking about shirts and laundry and how they fell. Since laundry doesn’t fall perfectly in position, the robot has to figure out in real time how to position the shirt. However, the repositioning can take quite a long time, perhaps up to seven minutes. Significant improvements are coming through machine learning, image recognition, and other AI components, helping make those robots better. We’re still a long way out from robots being able to do the simple task of folding a T-shirt.

I asked ChatGPT what question I should ask you. It suggested this: After a year of letting AI make decisions for you, what’s one decision you now refuse to outsource, and what does that reveal about where humans should draw the line?

It’s a great question. The example in the book is that I’m not going to let AI decide how I cut my hair. Every time I ask it, it comes up with a terrible idea. I don’t want to cut my hair like a bob, and I don’t want to cut my hair really short. So, I’m still not going to let AI tell me how to cut my hair.

Humans have agency. At the end of the year, when I asked the AI about something really crazy I could do, the machines suggested that I get a tattoo. I grappled with that because if I had agreed to everything AI had told me to do for the year, my life would have been destroyed. If AI had replied to every text and email message, I would’ve been divorced. I would’ve lost my job.

I had to set parameters so I didn’t destroy my life, so I could finish the book and be a functioning human and member of society. There are certain points where you start to realize—and I did as part of the experiment—that when you ask AI for certain advice, it really is a mirror.

AI only knows what it knows. It can only relay what it has seen or read on the internet. We are just not all that predictable. Sometimes we are. Sickness is a good example. It’s still so important to get a doctor’s opinion. Yet there were many times when it was quite right about the illness I had, or the rash from which my son was suffering. There were many times it was correct because of how it’s trained. But other things are just subjective, like haircuts. It does not know what my hair is like. It doesn’t know that I have hair, which, if cut really short, would look bad.

AI only has the pictures, but it doesn’t have the lived experience. Hair is a ridiculous example, but you can extend that to other parts of life and other parts of experience that only humans have.

What do you think business or tech leaders should be thinking about as they grapple with AI adoption?

Do we believe the “doomers” or do we believe in the utopian side that everything’s going to be OK? At the end of the book, I describe a complete dystopia or a complete utopia we could live in based on the [AI] life I lived over the past year and all the advice and interviews I did.

We will probably be somewhere in the middle, just like all technology. The response and the responsibility of the top AI leaders are so important right now.

There is massive competition between these models and tools [to determine which] will be the best . All I would want is a moment of pause from leaders when thinking through the bigger ecosystem. Of course, many of these companies do have good safety teams, and they are thinking about these things.

Regarding executives and the acceleration of this kind of technology in the workplace, I have advice for people leading organizations who are paying attention to this. [There’s a value to] knowing what’s happening on the ground, what’s happening in your organization, how people are using AI, and what it’s allowing them to do.

The narrative that “AI is going to replace us” is true for a certain number of jobs and tasks. But it’s also opening up so much more of what you can do. I just started a small business. My advice to everyone on my team of three is to tell me one cool thing they’re doing with AI each week and how it’s making the work better.

You discover you can use AI to do small things, and then that leads to more things. [Executives can] remain on the ground and talk to the people who are using AI in innovative ways and how AI is opening up new paths for them.

What do you want readers to take away from your book?

I really wanted this to be the AI book for the rest of us. There have been a lot of AI books out there—I have read most—and they are great. However, many of them are not accessible to a general audience. While I don’t want to pigeonhole certain types of audiences, it takes a lot to pick up a book and read a full book about AI or the drama happening in an AI company. The excitement is not there.

I wanted this book to be an AI book for everyone—for people who are questioning how this will fit into their lives. It’s for people who keep hearing AI is going to take jobs, that AI is going to be their doctor, that AI is going to be their lawyer. I wanted to answer that in a fun, lighthearted way, but that’s also very serious.

Visit Author Talks to see the full series.

Learn more