‘Betting on yourself is really hard.’ Joanna Stern on building New Things and why independence is worth it.
In her last column as a Wall Street Journal staffer, Joanna Stern, the paper’s personal technology columnist for 12 years, wrote what she called “a letter to my first-month-on-the-job self.”
“Allow me to introduce myself,” it began. “I’m future you — 2026 you, to be precise. And I’ve seen some things. Enough things to fill over 500 WSJ columns, nearly 450 videos and 167 newsletters.”
She also made “E-ternal: A Tech Quest to ‘Live’ Forever,” which won The Journal its first Emmy. The one story I remember, though, and the reason I first became aware of her, is her interview with Mira Murati, then OpenAI’s chief technology officer, about the company’s video generator, Sora. Stern wanted to know what it had been trained on. “We used publicly available data and licensed data,” Murati said.
“So, videos on YouTube?” Stern asked.
“I’m actually not sure about that.”
In March, with her book, I Am Not a Robot, on the way, Stern signed on as NBC News’ chief tech analyst and announced New Things, a newsletter and video company built around the question “Who is tech for?”
In this edition of Depth Perception, we speak with Stern about the sponsorship money she turns down (OpenAI’s included), why tech companies incorrectly expect creator journalists to be easier to deal with, why she made herself go to PC Magazine instead of ChatGPT to shop for a printer, and more. The following interview has been condensed and edited for length and clarity. —Parker Molloy
Learn from top longform journalists and find the best in-depth reporting. Subscribe to Depth Perception:
In April, you told Semafor that betting on yourself “feels horrible in the middle of the night,” that “it feels so terrible at 4:00 AM,” and that New Things had made “$0” so far. It’s seven months later. What does 4 a.m. feel like now?
Definitely better. We have definitely made more than $0, which is really good. Four a.m. would feel really bad still if we had made $0. I would be asking you for a job right now.
But I stand by that first quote, which is betting on yourself is really hard. This has been absolutely a journey. I’m having the time of my life and would not trade it for a second. We’re growing. But also, this morning I had a two-hour meeting with our bookkeeper, and I would say they’re kind of like a fractional CFO team now that are helping me figure out, what do I need for next year to grow? How do I sustain this? So I’m really enjoying all of that, but I can see why people would not want to do that if their main thing that they wanted to do is just do great journalism.
Before tech, why journalism at all? I saw an interview where you said you got hooked writing for, and then co-editing, your college newspaper. Was that it?
No. My mom started a public relations company when I was born. This was the ’80s. She was an entrepreneur herself. And growing up, I really wanted to be so much like my mom. I would go to her office, I would watch what she was doing, and I thought, “OK, I’m going to go into public relations. I’m going to take over the family business.”
So I did some interning there. I did interning at other PR companies. My summer jobs were working at PR companies, but during the year, I was getting a lot of journalism experience working at the newspaper, because I kind of thought, “OK, this will help with the PR thing.” But it flip-flopped, and the thing that I ended up really wanting to do was telling the stories.
In my first job out of college, I started at a PR company working on tech clients, one of them being Skype at the time. And I was reading a lot of tech press. I was reading columnists at The New York Times, at The Wall Street Journal. And I really was taken with [the idea that] they’re interviewing these people who are making these products. They’re giving their assessment of these products. And I just decided that’s what I wanted to do. I didn’t want to be pitching. I wanted to be actually writing those pieces.
I didn’t really formally ever go to journalism school, but I learned so much at every single job I had over the last two decades.
Depth Perception is brought to you by Long Lead, a journalism studio that partners with independent reporters to produce editorial features that inspire and delight. Over the past five years, our features have had a remarkable impact and their ongoing relevance has been unparalleled.
The goal of Long Lead is to produce journalism that is still valuable and memorable years after it is published. Our award-winning work has been cited by a judge ruling in favor of homeless veterans suing the government for housing, chronicled America’s gun violence epidemic so well that Harvard Law School made it part of its 2nd Amendment curriculum, and — at a critical moment in our country’s history — captured the testimonials of some of the last survivors of Japanese American incarceration.
Through all that Long Lead features have won or been recognized in more than 90 major awards competitions, beating newsrooms far larger and more established, time and time again. “The journalism on Long Lead is nothing short of stunning,” said NYU’s American Journalism Online Awards. “This is what I imagine long-form news will look like in the future.” See the future of journalism, today, at www.longlead.com.
You told Nieman Lab that one way you have to make money “goes so against everything I’ve ever learned at traditional outlets: I’ve got to know who the advertiser is.” Is there a company you won’t take sponsorship from? Have you turned anything down yet?
This is a full-on part of my day-to-day now. We have a sales company, Smooth Media, which brings me incoming opportunities, whatever big companies are looking for sponsorship right now. A lot of those opportunities, we say no to. I say no to a lot of things because it’s a conflict.
I will not take money from the Apples and the Microsofts of the world, at least in a way where I am out there doing the direct ad read. I think we’re going to be in a moment where we’re going to have to consider, “OK, are we going to do display [or] different types of advertising that wouldn’t require me to be involved with that as we grow.” But right now, I’m very firm about which kinds of companies I will do ad reads for.
We have absolutely left money on the table — lots of money on the table. In fact, it’s kind of a running joke in [our] sales Slack, where I’ll say, “Don’t tell me how much money we lost on that.” And the joke is, “Zero dollars, Joanna, don’t worry.”
I think the first, most important thing is to be transparent, because we are taking money from some tech companies. We have my Money Mic, and it’s very clear, the name of the sponsor is on it. We literally have a thing called the ad break, and you see it break the screen.
But I think this is getting to be just the way of the world for whatever you want to call our new genre, creator journalists. And [it’s] especially difficult for people covering tech, because that is where most of the money can come from.
So I’m very excited. We have one of our first CPG brands that we’re doing an integration with in November. I was never going to cover this company, right? Mercury was our launch partner. I was really never going to cover fintech or banking. But look, we’re ultimately going to have some tech companies in the mix. And we do already have tech companies in the mix.
Since going independent, you’ve done interviews with Mark Zuckerberg and Satya Nadella. Did anyone stop returning your calls once the Journal’s name came off your email?
Funny enough, it’s the opposite in the tech world. There’s a very big inclination from tech companies to want to work with creators, creator journalists, whatever we want to call ourselves. And I think that that’s twofold. One, they know that you don’t necessarily anymore have the baggage of, “I’ve got to go run this [through] 10 different editors” and all of that. I think they already can see that working with creators, smaller media companies, can be an easier process.
But also, some will say, “Oh, we really want to help you.” And I’m like, “Do you really want to help me? I don’t think you selfishly really want to help me. I think they see this potential for the ability to reach new audiences. And I think they think they have a little bit more control, too, of how that’s presented. And I think that’s probably not always right.
In my case, I think we’re already showing with some of the interviews we’ve done [that] we’re still asking tough questions. This is not a pay-for-play thing. Actually, this morning, we had something incoming from a smaller-ish startup, though they’ve raised billions of dollars. And I was like, we need to get on the phone with them because I think they’re expecting some really easy podcast interview. This company’s been under the microscope, and … we are first and foremost journalists here, and we’re going to be asking a lot of the questions that [have come up about] this company in the news. If they don’t want to participate in that, that’s fine. We’re not going to do a gotcha thing, which I learned really clearly at the Journal.
All my practices from the Journal are really still in place. Every story that goes out, call the company to let them know. We talk through what the positioning of that interview is going to be ahead of time. Yes, OK, you’ve got news, but we also are going to ask about all of the big news topics that are affecting your company right now. A great example is … we sat down with Greg Brockman, president of OpenAI. The interview was supposed to be about the future of computers, which it was. But it was that day that the Hugging Face [hack] happened. And of course I’m asking about this.
More from Depth Perception
Your bio says “professional asker of uncomfortable questions.” Is a question like the one you asked Murati harder to ask when a company like that could be a sponsor? Not that OpenAI would sponsor you.
Oh, 100%. OpenAI wants to be. I mean, I’ll be honest on the record here. That is one we’ve turned down. That is one our sales team has said, “What about them?” And it’s like, no. And again, [that’s] in the parameters of right now, where I’m doing ad reads. I just have to draw the lines, which makes sense for me. I can’t be in an interview with Sam Altman, or very soon, I hope, reviewing one of Sam Altman’s new products that is long anticipated, and say, “Oh yeah, back in November, we took—”
I can’t draw those lines perfectly for every company, but there are some companies [where it’s] pretty easy to draw those lines. And when we’re not able to draw those lines, [we’ll be] really transparent and say, “Hey, I’m covering so-and-so today, but actually they were a sponsor last month.” That hasn’t happened yet, but I can imagine it happening one day.
But back to your question, first of all, that wasn’t a very tough question. So I would 100 percent ask it again.
What made the Murati interview stand out was that you weren’t asking some trick question. It was just, how did you train this? And she just kind of froze.
There are a lot of tech podcasts out there getting access. If my access got cut off tomorrow, would it affect things? Sure. But the show is not all about interviews. This is not a podcast show that requires every CEO to come on and then want to come back. I hope they understand, “Yeah, she’s going to be tough and fair.” But if we don’t get you back, I still have a business to build. I’m not building this on access journalism.
You told Nieman Lab that by midyear, while writing your book, you “no longer needed the reporting assistant” because the chatbots had gotten so good. And your standards page says using AI is “a requirement to work here.” If entry-level jobs like research assistant start to fade, where does the next Joanna Stern come from? Or do you just not know? Because I don’t.
It’s tough, because, look, I hope to hire another entry-level journalist next year at this company. And I already hired one, and she’s amazing, and she’s learning so quickly. But she’s using AI all the time. Basic research, creating documents, templates: We use [AI for those] all the time. Even photo editing.
So just in watching Amaya, who’s our production assistant, go from production assistant to video journalist and editor, I think AI’s just fast-tracked that for her. But she’s also learning, I think, from me still so much about script editing and how we need to start a video, and where I’ll say we can’t green-light that story yet, because we don’t have enough reporting.
So I really hope that that’s what’s still happening within big newsrooms. That’s how I was working with my reporting assistants at the Journal. I think that’s how it’s going with interns. But the big question is, are we going to have less of them? Because we have less to do, because AI is doing whatever portion of that job.
But we’re using it a lot, too, in the reporting process. We did this piece on the Meta glasses [in June] where we wanted to go through Facebook Marketplace and look at all of the places that were offering the service to drill the recording light out of the Meta glasses. In the past, at the Journal, I would have had to gather a small team of people and ask the data science team, can you guys do a cohesive search of every state for these keywords? And they would be like, “OK, yeah, we’ll be back to you in two weeks.” You get the data in a report three weeks later, four weeks later, and then that story is going to be done two more months from now. That’s how investigative teams work at big places.
Here, we did that in like a day. We were like, “Hey, Claude, can you look through Facebook Marketplace, search these keywords?” It was making a spreadsheet, and it was pulling the links of all the places in the country. We went five listings per state. And then Amaya was going through and double-checking it, so we could go to Meta and say, “Hey, Meta, look, this is what we found.”
And I personally think that was one of the best lessons yet for her. She got to see how I wanted to approach a story, how I would previously have gone to a team, but now we went to AI to help with that. And yeah, we’re building a business, maybe people are going to be more forgiving about my use of AI because I don’t have the massive newsroom, but I think that’s the exciting part of this moment for entrepreneurs and journalists.
Your answer to, “Who is tech for?” is “you,” not “a very small group of people in a very small number of Zip codes.” You’ve said the most surprising thing about the book was how big a theme your kids became. Did the answer to “Who is tech for?” become “them?”
I get what you’re saying. I feel like maybe we need to be asking these executives this. But you know what they’re going to say.
I’m really happy you’ve asked the question, because I think it’s a challenge for me as a journalist and as an entrepreneur. Some of the most successful tech creators and new media outlets in tech are those that are really focused on the inside of Silicon Valley. You can make a lot of money and you can have a lot of power and influence if you just talk to those people. You don’t have to have millions of subscribers. If you have a very small, distinct niche group — which everyone’s saying right now is the future of media — you can squeeze a lot more money out of that, because these advertisers really want to hit those people. And you can also charge more on a subscription basis, because you know those people have money. So it’s a really tempting way to build the business.
I don’t only want to talk to those people. First of all, I think those people are really well served with the information they have right now. But also, who’s going to help the rest of the people? I want to still write in a way and explain in a way that’s really simple to normal people.
You told Semafor that some of the car time you used to spend on podcasts now goes to talking with ChatGPT or Claude, and that you get information that’s “not attributed to any specific outlet.” You’re building a business on people choosing your newsletter and your videos. What makes you hopeful about journalism?
I’m very scared about this. I’ve been reading the Anthropic S-1, and you read about the risks to the business. I always think, “OK, what are the risks to this business?” One, we’re so dependent on this YouTube algorithm right now. That’s not a great thing. My newsletter should be growing faster. But the sponsors really want YouTube.
And then another big risk would be this growing infrastructure of AI, and that’s how people get news. Even today, I’m looking to buy a printer. At one point I thought I was one of the world’s best printer reviewers and my first place to go was ChatGPT. And I was like, no. I purposely went to PC Magazine and I clicked on all their links. I didn’t buy it through them, but whatever. I gave them traffic. But I also was like, “OK, I know a human has written or gone behind these things to recommend the best printer for small businesses.”
And I thought to myself, I have to purposely steer myself, because it’s so easy to just go to ChatGPT to get that information. But I also remember how hard it is to test printers. That is a multi-day job to test one printer. So that is my biggest worry, that these things we spend a lot of time on just become one line in your LLM answer.
But in the same year that I have this fear, I also see this massive growth of our YouTube channel. The comments all say it’s so great to have a real human telling you, with great personality, what they trust and what they don’t trust. And so I think that ultimately goes back to why I got into journalism. I wanted to be the person that people would trust with the tech news. And we’re still sitting here in 2026, and that’s still true. Yes, the model’s different, and sadly, it’s all controlled by an algorithm we don’t control. But maybe there’s some hope in something I just said there.
Further reading and viewing from Joanna Stern:
- I Am Not a Robot: My Year Using AI to Do (Almost) Everything (HarperCollins, May 12, 2026)
- “Exclusive Mark Zuckerberg interview” (New Things, Sept. 23, 2026)
- “I paid someone to hack my Meta glasses” (New Things, June 3, 2026)
- “It’s Time. Meet My New Thing.” (New Things, April 22, 2026)
- “I’m Signing Off From This Column After 12 Years” (The Wall Street Journal, Feb. 6, 2026)
- “OpenAI Made Me Crazy Videos—Then the CTO Answered (Most of) My Questions” (The Wall Street Journal, March 13, 2024)