Atlantic CEO Nick Thompson on AI Deals, Creators and Media’s Trust Problem

0:00Hi, everyone. This is Scalable, a weekly show about the business of the creator economy. I'm Kaya Yurieff. And I'm Jasmine Enberg. Every Thursday, we take you beyond the headlines to tell you what's happening, what it means, and why it matters. The two most powerful people in the world are actively trying to destroy this industry. Even if we're all married to robots in three years, we will want to go to live events. Does social media die? I kind of feel like Twitter, TikTok, social media might just be destroyed.

0:31Hi, Jasmine. Hi, Kaya. And hi, everyone. Thank you so much for tuning into our weekly show about the business of the creator economy. You just heard the voice of Nick Thompson, the CEO of The Atlantic, and needless to say, one of the more candid executives in media right now. Before moving to the business side of journalism, Thompson was editor-in-chief of Wired, and he spent years as an editor at The New Yorker. He's also built a personal following of more than 1.6 million people on LinkedIn, where he posts near daily videos, making him somewhat of a creator himself. I was so surprised to learn that he films these videos in just one take because of course he has a day job.

1:06And he's also a competitive runner and even wrote a book about it called The Running Ground. So I really want to know if he thinks those robots that we might be married to in three years time are actually going to come with us to the live events that we're attending. Because could you imagine a room full of robots at Scalable Summit 2029? That is the stuff of nightmares. I have seen So many robots at tech conferences this year, and I've had enough, frankly. Anyway, back to business. Founded in 1857, the Atlantic has gone from losing money to being profitable in recent years and building a lucrative subscription business.

1:37The outlet is now expanding its newsroom, adding print editions, and putting on live events across the country, making it one of the real success stories in legacy media right now. One reason that I think they've been successful lately is that unlike most traditional news outlets, they've actually elevated their writers from day one. It's one of the few publications that have writers whose names I actually recognize and know and who I follow on social media, Charlie Warzel being the main one. And he really is kind of the face of The Atlantic for me. The Atlantic has also hired another well-known tech reporter, Will Orimus.

2:09And putting writers front and center like this is just so important as people increasingly want to hear from individuals rather than faceless institutions. On the flip side, this could obviously be a problem for The Atlantic as we're seeing more and more reporters go independent. Nick is very aware of that, and here's what he said. We think of our competitive status partly including Substack. We want Charlie and Will to write for The Atlantic and not to go to Substack. And so we have to figure out what are the things that make it valuable for an individual to be part of The Atlantic. That includes everything from editors to fact checkers to an established social media presence to lawyers.

2:45These are all things that independent writers, including us, don't have easy access to while established institutions do. He's also thinking about potential ways to bring more creators into the Atlantic. I would also like to create a creator network on YouTube, like find people who are amazing YouTube creators who have Atlantic values and standards, who aren't making enough money on YouTube by themselves that they can sustain it. I sat down with Nick at a tech conference in Paris last month, and the conversation covered a lot of ground, as you can see. We talked about everything from where he sees the media industry going to the Atlantic's deal with open AI and whether AI chatbots owe publishers a cut.

3:18He also gives an updated sense of the Atlantic's financials, and we debated whether social media as we know it survives in this new era of AI. Let's dive in. Hi, Nick. Welcome to Scalable. Thank you for being here. Thanks for having me here. The Atlantic has positioned itself as a writer's collective, and so many of your writers have made names for themselves. Charlie Wurzel is someone I followed for a long time. Will O 'Ramus. And the media industry, obviously, has been in major upheaval. More reporters are going independent. There's AI, which we'll talk about more. More people are getting their news from social media and creators.

3:49What has all of this meant big picture for your business? Well, it's super interesting, right? And we think of our competitive set as partly including Substack. We want Charlie and Will to write for The Atlantic and not to go to Substack. So we need to give them more as part of this writer's collective than they get from Substack as a writer's collective. And so we have to figure out what are the things that make it valuable for an individual to be part of The Atlantic. We need to make sure what are the things that make the collective valuable to the reader and what are the things that make the collective valuable to the writers, to the editors, the fact checkers, the copy editors.

4:22So it's a totally different competitive dynamic than we had 10, 15 years ago. So what does that entail? I should preface this by saying I'm on the business side, not on the editorial side. But you've been a journalist for a long time. I don't even talk to Will. I have not talked to Will once. Oh, no. I slacked him once. I was like, congratulations, welcome to The Atlantic. I deal with the advertisers. But if you're a person like them, right, and they have large social media. And Charlie even had a sub stack and left media and came back. Came back to the Atlantic. So we give a lot of things. So you get fantastic editing.

4:52You get fact checkers. You have lawyers who review your pieces. You have the best copy editors in the world. You have the social media power of the Atlantic. You have the Atlantic's website, which has a URL with good SEO. So there are all of these things to come together. I think the most important is the writer-editor relationship. And I've been on both sides of that. I worked as an editor for a long time at The New Yorker, where my whole job, my whole life was help these eight people get their best ideas into the world. And when you start a Substack or you go independent, you don't necessarily have that.

5:26Very few people on Substack have editors who are fully devoted to that. Maybe they do, or maybe they're married to someone who reads their stuff. But to have an organization like The Atlantic with the senior editors is super valuable. Now, it's obviously not enough for everyone. We do occasionally lose people who go and become creators, but we don't lose a lot of them. Do you ever think about hiring creators as writers or some of those sub-stack? We've seen the Washington Post launch a creator network. CNN just hired a financial creator, Kyla Scanlon as a contributor, an analyst, which is super interesting.

5:59Fast Company has experimented with sub-stack writers writing original columns, but what do you think about that? Yeah, definitely. I mean, we did at one point, right? We started what was called the Atlanta Subscriber and newsletter program. And that's actually where Charlie Wurzel came in. And we hired six to eight people who had their own newsletters, their own personal newsletters. You know, Zochie Gonzalez, Molly Jong-Fast, Charlie Wurzel came in. And some stayed, some left. We ran it for about a year. And the idea was, we'll have them come in. We'll have them bring their audiences. And then when people subscribe to The Atlantic, they'll get access to the newsletters.

6:30It was kind of trying to figure out a win-win mechanism. When was this? What year? This would have been in 2021. Okay. So early. And then the economics didn't really work. Or maybe the other way to put it is I couldn't figure out the economics. I couldn't figure out a way that made it work for everybody. Like we weren't able to generate enough subscriptions to justify the salaries. And so we switched them to just write stories for The Atlantic and run their newsletters. So we have tried it. We certainly could try it again. We have nothing against creators. I mean, anybody who can bring their own audience is amazing, right?

6:59That's great. Particularly if they have an audience that's different from The Atlantic. If their audiences, you know, they have 12 ,000 subscribers and all Atlantic subscribers, there's no marginal value. If they have 50 ,000 subscribers and they're actually not Atlantic subscribers and bringing the creator on gets them to read the rest of the Atlantic, that's amazing. You know, those are mostly decisions for the editorial side, but yes, totally in favor of it. And also, where do the smartest young journalists go? Like there used to be a path, right? Yeah. You would go to school, you would work on your school paper, and then you would go work for a local paper, right?

7:24You would go work for the Concord Monitor and then you would move up. You'd work for the Boston Globe, you'd work for the New York Times, and you'd work for the Atlantic. There's like a ladder that you would climb. Now what are those steps, right? is the Concord Monitor still hiring? I don't know, right? The Boston Globe's doing well. Time's doing great. But like local news, because some of the rungs of the ladder have either been like knocked out or kind of faltering, and because Substack exists, like a lot of smart people, you go for your school paper, now what do you do? You start in Substack, right?

7:50You start like writing on Instagram. Or you work for someone else's Substack. Yeah, right. And so I think that there are going to be a lot of people who, folks who are writing features through The Atlantic in five years, some of them are going to have to come in through interesting routes like creators. I would also like to create a creator network on YouTube. Like find people who are amazing YouTube creators who have Atlantic values and standards, who aren't making enough money on YouTube by themselves that they can sustain it. But if they were part of an Atlantic network, they could make more money and then be good for everybody.

8:20When you say people who aren't Atlantic subscribers, are there certain demographics or types of people that you want to reach more than you are now, who creators could help with potentially? Like everybody, we want to get younger, right? Our subscriber base is younger than you, Mike. think. But in our average, we are still in their 40s. We're more conservative than you might think. It's like 35%. 30 % identify themselves as conservative as opposed to moderate or liberal, which is pretty interesting. But we want to reach the whole country. There's no particular demographic group that we don't reach.

8:48But if someone reaches heavily into a demographic group or more, that'd be amazing. Are you very US focused? How big is your international? Our international, there's about 10 % of our subscribers. I would like it to be more like 15. and we're actively looking at different markets. And in fact, it might be a way to get into new markets. Like, how can we get into India, right? But whether you push hard in a new country depends on lots of things, right? Are there a lot of English speakers? What does the news ecosystem look like? Is there a culture of subscribing to magazines, right? Like some places you just buy one-offs, right?

9:18And there's, you know, you can't really, our business wouldn't really work or it's very hard to get people to renew subscriptions, right? So you have to look at all of those economics. So, you know, we're looking at Australia, we're looking at Canada, we're looking at India. Yeah. One thing the Trump administration has done is it has certainly increased interest in reading The Atlantic among people in Canada. Why do you think that is? Don't know. Something about the 51st state. So The Atlantic has been expanding. You're hiring more reporters. Yeah. Profitable. Yes, very weirdly. Exciting. What can you tell us about how the business has been doing?

9:48There's been figures floating around. The latest I saw was from Wall Street Journal a couple of years ago that put your annual revenue at about$100 million. You can do the math. We have 1.5 million subscribers and we charge about$90 a subscription. throw in a bunch of advertising revenue. And it's not hard to figure out exactly how much money we make, but it's more than$100 million. What's gone well is we publish amazing journalism. We hire writers who people want to read. You name a whole bunch of them. You associate them with The Atlantic. It makes you feel good about The Atlantic because you feel good about those people.

10:15And then we built a really, really effective subscription mechanism. And we did that through a lot of testing, a lot of work, a lot of product iteration. So what should the paywall be? What should the paywall be for people coming from different referral sources? What should the price be? What should the price of the bundle be? What should our paid marketing budget be? What should we set the CPA at? What color should the paywall selection page be? Do you want red? Do you want blue? Do you want thick red? So you run all of these different tests. And there's almost an infinite number of variables that you could adjust.

10:48So one of the most interesting ones, which we haven't solved, is how much of the story do you show someone before they hit the paywall? Should you show a summary of it? Should you show quotes from it? Should you show people's tweets about it? Or should you just put the first 200 words of text, the first 100 words of text, the first 600 words of text? If you put more text, you can show more ads, which bring you revenue. People get in, but then maybe it's a substitute. Maybe they don't feel like they need to subscribe to read the whole thing. If you put 100, well, A, you can't really have ads. And B, do they get enough of a taste in it?

11:15And so those are really hard questions. But they're the kind of questions where if you can do 1 % better, you actually can change the trajectory of your business. And so over the last four years, what we've done is we've committed ourselves to try to identify what are the most important leverage points. Often you get clear data. Sometimes you don't get clear data, so you just go on instinct. Sometimes your instincts are wrong. Sometimes your instincts are right. So we're constantly running tests and iterating. And through that process combined with the great journalism and the great writers, the business has gone from losing a lot of money to making money.

11:50And how big of a revenue driver are events? You're doing these 50-city tour over two years, and obviously subscriptions have been growing. I mean, is that sort of the future? So our events business is good. I think it's a much smaller percentage of our total revenue than Semaphore. Semaphore is really an events business, right? And they're leaning at heart, and they're doing very different events from what we do. But we are going to go to every 50 states, each of the 50 states, over three years, not two. I don't want my events team to do 25 events a year. 16 is hard enough. The goal is to get to every state by the end of 2028.

12:22Okay. And we're not going to drive a ton of money or a ton of subscriptions from that. We're going to make money on it. So why do it? Well, because we're a magazine about America, right? You should be everywhere. I hate that the political conversation in America centers in like eight states because they're swing states. Everybody goes to Pennsylvania and all the candidates are in Pennsylvania. Well, I don't know. They should be in Massachusetts. They should be in Montana. They should be in Utah. They should be in Texas. And The Atlantic is a magazine about America. And it was founded by people who wanted to prevent an American civil war in 1857.

12:55And it's our obligation to go everywhere. And we need to go to red states. We need to go to blue states. We need to talk to people in red areas. We need to talk to people in blue areas. And I just want us to do it for the sake of the Atlantic. We didn't set this program up because I was like, let's figure out how we can make the most money with events. Because to make the most money with events, what you would do is you would do a big event on the future of healthcare or something where there's like lots of sponsor interest where you can get tons of advertisers. You would do it in a big stadium.

13:26Like right now we're in an event space. It took me like an hour to get in to Viva Tech, right? It took me like another hour to find you. So our events in, you know, Bozeman are not like this. We're doing them to go and to reach people to find new ideas and to go to new places. And I'll also add this to sort of the second part of your question about where the future is going. I don't know what's going to happen to print journalism. Although you added two print editions. I did. And we can talk about why because it's kind of interesting. AI is changing so much so fast. And it's changing the way people read.

13:57And it's changing the way people write. And it's changing the way people find stories. And it's changing each of those three variables, which are the core three variables of our business. And it's like spinning them up, turning them inside out. And there are all kinds of features for all of them. But one thing I have high confidence in is that, you know, even if we're all married to robots in three years, we will want to go to live events. Actually, most Atlantic people won't be married to robots. But even if like 70 % of America is married to robots, there will be a lot of people who want to come to Atlantic live events and see David Frum and Ann Applebaum and Willa Ramis in person.

14:31And I think that the AI-ification of society is going to put a premium on live events. So I want us to be more ambitious. us, I want us to be more aggressive about how we pursue our live strategy. So there's a new Reuters study that just came out that found that only one in four Americans trust the news, quote, most of the time. So we're hitting record low on trust. I sort of can't believe I have to ask this question as someone who's worked in traditional media for the last 10 years, but why do institutions and legacy media companies still matter if you have this trust gap in the U.S.? There's two great questions.

15:02There's also, why is the trust declined? Yeah. Well, they still matter because there is plenty of trust, right? There are people who don't trust us and we need to win those back. And going to all 50 states is part of the strategy to try to... I would like that number to stop going down. Right. I would like it to at least go down slower. I would like that number to go up, but to get to go up, you need it to go down. You need to get to stop going down so fast. They still matter because they have huge reach because they're important and because the people who do read them trust them immensely. One of the big troubles is that the business model that works for places like the Atlantic is a paywall.

15:32And necessarily with the paywall, you have fewer readers. And so you can't reach everybody. You can't put your fact-checked, legally checked, heavily reported stories in front of everybody. And so we've kind of created an information ecosystem where bad information is free and a lot of the good information is paywalled. And that's a structural problem that I don't know how to solve. And I can't solve it by lifting the Atlantic's paywall because then I drive the Atlantic out of business. To your question, they still matter because they're hugely influential, because lots of people read them, because lots of influential people read them, and because we still have many millions of people who read us.

16:02And a lot of the political conversation, cultural conversation is set by publications like this. So they still do matter. The challenge is how to win back trust. Why do you think trust is so low in media? And like you said, it's been going down more rapidly than you'd like. What are the biggest factors driving that? The two most powerful people in the world are actively trying to destroy this industry, Donald Trump and Elon Musk, both because they just don't want anybody to hold them to account. And they've had stories that have been run about them that they don't think are fair, they don't think are accurate.

16:30And so one of whom owns, you know, one of the central town squares and the other of whom is in the White House. And so the constant weaponization or the constant use of those platforms to try to crush the reputation of media has been somewhat effective. So that's part of it. Part of it is it's just a dispersal. Right. And so they don't trust media, but they do trust the subtext that they read. Right. And so they don't trust media as a whole. because Trump told them not to trust media as a whole or they're just generally untrustworthy. So part of it is like, we all hate Congress, but we like our Congress people.

17:07We all hate media, but we like our media. People trust people. People trust people. Some of that is our fault. And some of it's our fault because of the way the incentives are set up, right? And so incentives that push people towards clickbait journalism, that push people towards curiosity gap headlines, that pushed reporters to say salacious, crazy things on social media. because you can torch the reputation of a whole newspaper if you have one reporter making one bad tweet. One reporter making one good tweet really doesn't get seen. One reporter making one bad tweet can hurt reputation a lot more.

17:37There are many more ways. There are a whole bunch of factors at play. And also just trusting institutions has gone down across the United States. But it has gone down more in media, so I can't use that as an escape outlet. So some of it's our fault. Some of it's, you know, powerful enemies. Country works better when media acts in ways that wins trust and when people do trust media. So let's talk about AI. So many publishers are rightfully worried. The Atlantic was one of the first to strike a deal with OpenAI. Why? Because we were worried. We would prefer it if people didn't steal all of our content.

18:04If they do steal our content, we prefer to be compensated. You have two robbers. One robs your house and doesn't offer you any money back as an apology. And the other robs your house and does offer you money back as an apology. The whole notion that you should prefer the robber who doesn't offer you anything back is a strange part of my business. But there we are. Anyway, with OpenAI, all the AI companies came and they took all our data. They believe it's part of fair. They believe it's justified under fair use. We don't. that is going through the court system. We're part of some litigation. Others are part of other litigation.

18:30In any case, some of them came to us afterwards and were like, okay, actually, we would like continuing access to your data and we'll pay for it. We'll license it. We won't continue to steal it. We're like, great. We'll do that. And we'll, you know, you can pay us. And then the most important part of it was we knew that they are building kind of the future of search on the internet and that people are going to find information through OpenAI and through ChatGPT. And we wanted to be at the table talking with them as they built it, right? We wanted to make sure that they would build a search engine, an AI search engine that wouldn't reproduce more than a certain amount of stories, that wouldn't let you write in the voice of one of our Atlantic writers.

19:06We wanted to be able to help influence in that product. And so the opportunity to both be paid and have influence and presumably get a little more search traffic versus none of those things seemed like obviously you should take that deal. If you think you can get more by suing them and get more leverage, sue them. But if the choice is do nothing or take a really good deal, I always decide to take a really good deal. And they've been terrific partners. They've been great to work with. They've been very helpful. AI search has not led to as much traffic as I would have anticipated, but certainly there's been terrific communication between the companies.

19:40Why do you think that is? People like to just stay inside the search interface. How often do you click on a link out of ChatGPT or Cloud or Gemini? You don't. right? And I thought people would, right? Because we learned this habit. You go to Google, you say, hey, what's going on? Google says, this is what's going on. Here are some links. And you, oh, cool. You click on the link and then you're in a different world. With the chatbots, hey, what's going on? Here's what's going on. Would you like to tell me a little bit more? Sure. Hey, here's the answer to that. How about this? Oh, great. Let's stay here.

20:07And so you just stay in the chatbot. And so the old ecosystem, Google was changing over time to keep you on their page more and more, but they sent you off. And that's not the way this new ecosystem works. It's not the way people want to use it. It's not the way it's being built, unfortunately. Now, maybe that runs against what I initially said, which is that we've had these great conversations. Because obviously, I would have liked ChatGPT to have built in such a way where, hey, what's happening in Iran today? Well, it looks like there's a peace deal in the works. Here's a great story in the Atlantic.

20:34If you'd like to read more, read this story in the Atlantic. Instead, what they say is, hey, there's a peace deal in the works, right? Here are the details of it. And then there's a little link to the Atlantic that nobody clicks on. So what should this look like and what makes a good deal? The most important question, this is by far the most important thing for a business, is the what should it look like. And what it should look like is different. And what it should do and what I would like is for the chatbots to recognize who is actually someone who's trying to get to the Atlantic or who wants the depth that the Atlantic can provide and who wants the relationship with the Atlantic readers.

21:04Like if they search, like, what did the Atlantic say about Iran? Or there's something about those queries where they've come in a few times and they've asked about things that we've published about that day. then the search surface should send them to the Atlantic, right? And they should recognize that the person is going to have a better web experience on the Atlantic instead of inside of Gemini or Grok or wherever it is. And so there's a certain set of queries and a certain kind of people where they will have a better experience if they stay inside the chatbot. And there's a certain sort of people in a certain set of queries where they'll have a better experience if they go to the media entity.

21:33And what you would like is instead of having the AI bots try to absorb everything, to send that second group over. If Gemini is rebuilt in such a way that they're able to use the power of Google to identify who really should go listen to the whole conversation on the scalable pod, as opposed to just these little summaries of what Nick said. Let's send that person over there, and let's actively send them over. And if the way Gemini evolves is just two people at Google, and they're arguing, and one is like, you know, it's better for us to just like everything. We'll scrape the data of the scalable pod.

22:03We'll summarize it. It's got to keep them inside Gemini no matter what. And the second person is like, no, I don't think so. We're for the user. Some of them we should just summarize it. Some of them we should send them over to listen to the scalable pod, both because they'll have a better experience and because we know it'll be better if the scalable pod survives and has listeners and advertisers. And it depends which person wins that conversation in the debate at Google. And so my job is to help the second person and show that it's better for Google if people are going off into the media ecosystem, both because the media ecosystem survives, and Google needs the media ecosystem to survive in order to have great content fed into it, and also because it's better for the user.

22:38Yeah, because that was my question, Because for the AI company, I can understand why they want to keep people in their ecosystem. But they also, you've said this in the past, like they can't run the media business out of business. Because then what are they going to have on their chatbots, right? They drive you out of business, right? This conversation is good for Gemini. We're creating a bunch of information that goes into Gemini and makes lots of Gemini answers better. If Google destroys your business model by eating everything, well, then that's not good because the information doesn't flow in.

23:03And so I have to use whatever knowledge, skills, data I have to help work with Google, OpenAI, Anthropic, Proc, Facebook, everybody. So you said on a podcast last year that AI is going to change every single thing about the media industry. So what are the biggest things? You can just kind of go through the whole chain of what a story is, how people read it, how it's monetized, and look at the way. So what a story is, right? So AI is going to change the way we report, right, and find stories. We have a rule at the Atlantic that you can't use it to write a word, which I think is correct. If a person's name is on the story, it should be all their text.

23:38But if you're a reporter and you're not using, you're building an agent network to find stories and analyze stories, you're going to be outcompeting. And I'm not a reporter anymore, but to the extent I work in these models, it's incredible. Okay, so that's stat one. Then the story is edited and copy edited. Like, obviously, AI is super useful. Distribution AI is super useful. Where should we distribute it? How should we feed it into the paid marketing engine? So all of that is really important. Okay, so now how is it read? So AI is going to change the way people read. Like it used to be, you'd go to your website, you'd open your magazine.

24:07Now you're going to send that to your agents. You're going to say, hey, like summarize the news for me. Find what's interesting to me. And you may not even read it. You may listen to it. You may, Lord knows in five years, you may just absorb it from like, you know, vibrations in your wrist. So AI is going to, probably not that, but like we're going to be like wearing devices and we're not going to be holding magazines. or like people under 30 are going to be holding magazines. And so AI is going to change like the consumption of it. Then you go to, you sort of jump over to the business side. Okay, so now what are we going to do about subscribers?

24:40Like people aren't buying the magazine. They're not coming to our website. How are we going to get them? They're sending out agents. Are we going to make the agent subscribe? Are we going to make a deal with a network of agents? How are we going to get advertising into it? Okay, so L 'Oreal wants to advertise in the Atlantic. Great, thank you. But how do we get information about L 'Oreal to the agent? What is the transaction there? So it makes a huge difference there. And then there's like the licensing side where we're making, you know, we're going to make, we have a partnership with OpenAI. We have a lawsuit against other companies.

25:07We are in negotiations with others. We've licensed our data to other companies. So there's that whole element of it too. And then I also think they're going to be AI native startups that compete with us right now. Like, you know, when the web came around, I mean, first thing it did, it starts to like change the business model of the incumbents. And then suddenly it creates this whole new ecosystem, right? It creates a business model for Vox and BuzzFeed. I worked at The New Yorker from 2010 to 2016. Who is The New Yorker competing against then, right? It's competing against a bunch of publications that had not existed in like 2005, right?

25:37Turned out that weirdly The New Yorker won because the digital startups kind of rode a wave that then crested. Well, they were so reliant on the social platforms. And The New Yorker, like to its great credit, was not as reliant in part because it had like sort of a long-term infrastructure built up in part for other reasons. But who knows what's going to happen, right? There's going to be another wave of AI-native startups that we're going to compete with, and some of them will be incredible. So it kind of changes every single thing we do. And then, of course, we're a business. So how much of our customer service will be human?

26:08How much of it will be AI? What kind of new products are we going to do? We've just built a ticket tool for our engineering team so they can just say, hey, the alt image feature on Android devices isn't working. Can you just fix that, Claude? And so ideally, it'll make that process much more efficient. So you sound quite optimistic, actually. I think there's so much fear, but there's also excitement, right? Like this is a very transformational technology at the same time too. Yeah, of course. I mean, everybody's terrified. My business is just like, they're just so scared. They're scared into paralysis, a lot of people.

26:38And sure, it's scary. Sure, there's going to be new competition. Sure, our business model is like, it's talking really about how fantastic our business model is. Will it still be like that in two years? I don't know. But the worst thing we can do is say, oh my God, AI is scary. and therefore just like stand out there and be like, bad AI, bad AI, and then like not do anything. What we should actually do is experiment with all the tools, like study what it's going to do, try to harden our business model, figure out how we can evolve. So that's what I'm trying to do. And then how does events and then bringing back more print editions fit in with that?

27:11I missed a crucial step when I was explaining, you know, Google goes away, search goes away, like far fewer people come to the Atlantic. Okay, so what we need to do is we need to figure out as many ways that we can directly get to humans, that we can take the Atlantic and get it to a human without any AI in the middle, because the AI is unpredictable and weird. I want to win in every way I can with this unpredictable and weird AI middle, and I want to avoid it as much as I can as well. So more print magazines, right? That just goes to the US Postal Service. Not a lot of AI there. Live events. We sell you a ticket.

27:44Maybe we market to you with AI, but we have a human on a stage in a chair, human in the audience in a chair. Wonderful. Newsletters. Again, there's some AI distribution mechanisms in there, but we don't have to go through a social platform like Facebook or Google or Twitter. You sign up for our newsletter, we send you the newsletter. We have an app. You sign up for our app, you go to our app. The closer I can get the Atlantic to you, the better off I am. That's why we increase the number of print issues. That's why we're increasing the number of live events. That's why I've got a bunch of engineers really trying to fix our anything that's wrong with the app, that's why you've put a lot of work into newsletters.

28:18So you have to fire on all cylinders. Yeah. The other question too is what's a good deal look like? Now the citations are better, but creators and publishers are still not getting paid for that. Should they be? For citations? Yeah. Should there be some sort of revenue share like there is on YouTube for creators? When an ad runs on your video, you get money. Is there something like that for chatbots that makes sense? Yeah. So the model, the ethical just model is what ProRata has built. For those people listening who don't know Prorata. What they did is they said, okay, everybody opt in and we will run a reverse algorithm and we will determine how much your answer contributed to the answer that a user gives.

28:57And so if the user gives, asks a question, like I was just going into my chat bot and I was like, hey, I need to figure out this thing. I'm interviewing Tim Berners-Lee in a couple of hours and I don't understand this thing about what he's building. And Claude answered. Now, Claude answered not because Claude knows, but because Claude read CNET, Claude read The Verge, Cloud Red Wired. And so in the ethical version, which is not the version Anthropic has, despite being known as the ethical AI company, they would run the algorithm backwards and they would say, okay, so Nick's answer was 17 % from the Verge, 23 % from CNET, whatever from Wired.

29:28This answer is a certain percentage amount of money that Nick pays Anthropic and there's a rep share on it. And then, so we will give, you know, a penny to the Verge and a penny to CNET and kind of like how Apple News works. And they send the money out and they apportion it based on what you read. That's how the AI bot should work. That's what ProRata built. That's not how they built. The way they do it is they just take, and they take your money, and they keep all your money, and they don't pay anybody. So you're in favor of the rev share. Yeah, the rev share, obviously. And it's criminal that they haven't done it.

29:59And it's going to have to be forced either through the courts or through legislation or through continued pressure from our industry. I wish a company like ProRata was winning. ProRata is an amazing company. It may yet win in some way, but like a lot more of your listeners are using Claude right now. I wish Claude had been more ethical about this. They've been so interesting in their ethics in so many ways, right? Like their constitution. Oh, wow. So interesting. Like the way they launched Fable, like obviously led to a catastrophe, but wow, huh? We'll nerf it if you ask a hard question. No, that's kind of an interesting thought, right?

30:36Like instead of we'll deny the question, we'll just like nerf the model. Interesting, right? they've done a lot of really creative stuff. What they have not done is compensated anybody who contributed to their answers, which is a huge sin. So they need to fix that. Okay. To pivot a little bit. So you have a huge personal LinkedIn following, more than 1.6 million followers. And you post a lot of really interesting videos as well. That's a big part of your strategy. So first, why LinkedIn? Why not do this on TikTok? I do do it on TikTok. Oh, you do? Okay. But LinkedIn's the biggest one. Isn't that your focus?

31:07No, it is my focus. Here's the question. Yeah. Why do people on LinkedIn like it more than people on TikTok, right? Well, I think it's a subject matter, no? So eight years ago, I was like, you know, I kind of want to figure out more how social video works, vertical video works. You know, I was then the editor-in-chief, so I didn't get to report very much. I wanted a mechanism that kind of forced me to follow the news and summarize the news. I was like, you know what I'm going to do? I'm just going to like talk about the most interesting thing in tech every day. And so day one, I'm like, okay, I'll take my camera, I'll balance it next to this water cup, and I'll talk into it.

31:35and I now have done that pretty much every day for eight years. And at first, it wasn't like I intended it to be on LinkedIn. I put it on YouTube and the algorithm was like, this kind of like skinny journalist is like, who cares, right? I put it on Twitter and people are like, gain some weight. You put it on LinkedIn and people are like, oh, wow, the editor of Wired is talking to us about like, you know, neural networks. That's amazing. And so I would experiment through the year and I'd be like, okay, I'll try again on Twitter. So you still do this every day? I still do it every day. I mean, not every every day, but certainly four days a week, usually five days a week.

32:09I did one on Saturday, this Saturday when the anthropic thing happened. Oh, that one. You had a good hook. You said the worst AI regulation idea I've ever heard. Yeah, well, I was - That's a TikTok of you. Well, I was pissed off. I usually, I try to be like, I'm the CEO. I'm not supposed to have opinions. This is like on the one hand, on the other hand, right? Because you work with advertisers, you journalists, like you don't want to bias things. But occasionally I get really mad. And the notion that the federal government is going to stop foreign nationals from working on it, literally the best thing about America is that we're a place where people come and build things and they want to come to America because they get free speech and they get great institutions.

32:43They can work for companies with venture capital. They can build things. That's the best thing about our country and our economy. Besides the White Mountains, which are amazing too. God, I love hiking up there. So this amazing country has this amazing thing that builds its economic strength. And the Trump administration is like, we don't like this. Let's just make it so no foreigner can work at an AI company. Why would you do that of all the like terrible ideas? So I got very upset. So occasionally I'll get upset and those things tend to do better. But sometimes it's the best one is where I'm just like, this is what happened.

33:14I'm just like summarizing or, you know, hey, there's a new AI paper out. It's kind of weird. This is what it says is the implications. These are the flaws. I do that every day and I experiment over time and it just worked better on LinkedIn. And I started getting this huge following on LinkedIn. so I kept doing it. But it wasn't like I intended it for LinkedIn. I was still much more of a Twitter person back then. I don't know why I didn't start putting it on Instagram, but Instagram turned into just a place where I'd talk about running and put pictures about my races. So it really just popped on LinkedIn.

33:42I now do it on TikTok. It's starting to get engagement on TikTok. We're now in the five-figure followers, not the seven-figure. So getting there, YouTube is picking up. We're now in the five figures on YouTube. It's exciting. It's hard to grow on YouTube. It's very hard to grow on YouTube, but it's growing well. How much time does it take you to do these videos? I do them all one take. Do you script it? No. What's your process? No. I just go. Which, A, I think it adds to like the, it feels real. I missaid something yesterday when I was taping, right? I can't remember exactly what I said, but I like, I kind of had to double back and repeat the word.

34:13And if you're trying to do a clean take, you would go and do it again. But I like, I'm, I've got a job. You're a yo-loi. Yeah. I'm CEO. I've got to like. You don't have time to be doing. I can't. Because once you start doing multiple takes, then like why do two? Why not 10, right? If you do 10, why not 20? Like can you really get the person? Oh my God, my hair is sticking up. So it's really, it's just like I have my briefcase and in it I have like a little mini tripod and I have a microphone. And if something great comes up today, I'll do it. Is it helping you connect with people? Do you feel like it helps people connect with the Atlantic more?

34:44Like what have been some of the benefits? Yeah, it helps people connect with the Atlantic. I think there's a subset of people who are like, I don't, you know, the CEO, he's that guy who talks on LinkedIn. It's very helpful for advertisers. You know, advertisers are all on LinkedIn. The Atlantic's advertisers, like we have a lot of, they're like thought leaders. Four of them are your biggest ones. So our biggest advertisers, you know, they range from Hauser & Wirth to JPMC to PwC has been a big sponsor of my podcast. Biggest advertisers this year are TIAA, Visit Toronto. So they're like, they're places that are appealing to influential, tech-fluent people.

35:19And you have decision makers, right? TIAA wants like decision makers, right? They read the Atlantic. And so to what degree does a company like that know about the Atlantic? Because of my LinkedIn. To what degree does it help? Small degree. There's also this other business where I host events and I host dinners constantly for advertisers. And that also benefits from my video series. So I didn't set it up to do that. The other benefit, this is maybe the interesting benefit. So I have a job where it would be very easy for me to spend a whole day just, okay wait this person asked for a raise and we're going to start the fellows program over here and let me check in on the product thing here and let me push this here in the paid marketing division and there's like a gap between the engineer and the designer let me see if i can you know sort it out and actually can we like get this a little more data about whether we should expand it in you can do the whole day running that and that's great and that's how i spend most of my day and i love it right and it's the most important thing is that the Atlantic succeed but if you do that then you're kind of missing.

36:19There's no moment where you're, you could say, I'm going to spend the first 30 minutes of my day reading the AI news, but you probably wouldn't, right? Because the other stuff would bleed into those 30 minutes. But because I have this forcing function, I'm actually going to have to say something by the end of the day. In fact, people expect me to say something. It's kind of a forcing function to like be reading interesting AI papers and to be following the AI news and to like the government shut down fable. What? All right, let me try to figure that out so I can explain it to people. Not let me just like read the New York Times story and kind of absorb a little bit of it, but let me try to get to the guts of it because I'm going to explain it.

36:49And then that's super useful because I have a much better understanding of what happened with Fable. If I were a better person, I could just like learn without having to present on LinkedIn every day, but it helps me learn. Well, it keeps you accountable a little bit. Yeah, totally. Do you feel like your running ties in with that too? I mean, you run to work every day. You're an avid runner. I mean, we could all just work all day, but having these other things going on, does that help? It's nice. I mean, the running is useful because I run in the morning, I run at the end of the day. And it kind of like allows me.

37:17So I wake up, I work, then my kids wake up. I play with my kids, then I go to work, then I run home, I reset, and then I play with my kids, and then I work again, right? And you make your LinkedIn videos. And I make my LinkedIn videos. And so the day has kind of those different blocks, and the running is like sort of separates them and allows me to like clear my head. And there are all kinds of things about running that help with work. In fact, I wrote a whole book about it. It is mostly a way to sort of lock in meditation. All right. You've hinted throughout this conversation about kind of where things are going.

37:46You have the rise of the creator economy of AI. It's harder to know what's real and what's not. Like, where is this all going? I almost wonder if we boomerang back to then trusting institutions because there's just so much noise. It's going to be both. It's like a barbell, right? And there will be, it's not quite like a barbell, it's like a barbell with, I don't know, like with three weights. I don't know, metaphor is not working, but there will be a boomerang back to trust, right? And so institutions that can hold on to trust are going to become ever more valuable, right? Because everybody will be able to create information.

38:16There's going to be so much stuff on the web and you're going to have to know what to believe. Now, holding onto that trust will become harder, right? Because the Atlantic could torch its reputation tomorrow if we, I don't know, accidentally, why don't we want to say how we could do it? Because then someone will figure out. But you know, there are people, Dinesh D'Souza, right-wing and intellectual prominent Twitter following him, about every six months, he will tweet out a fake Atlantic story, right? It'll be like someone will have doctored a headline, maybe once a year. And it'll be a doctored story that says the Atlantic said this crazy thing.

38:49And it'll be like, can't trust the mainstream media anymore. And you're like, okay. And so then we go, Dinesh, this is not actually an Atlantic headline. He's like, well, it could have been. But then he'll delete it. But he's doing this. This is just pure misinformation. He's not fueling it. He just saw it and thought it was real and is posting it. But you get to these nodes, right, that send out the disinformation everywhere. It's very hard to shut down. With like Dinesh, you can call Dinesh, right? He has a phone number and email and you say, Dinesh, this is false. And Dinesh will take it down.

39:13But like it is then by that point, you know, what is the line? The lie gets out before the truth gets its boots on. that's going to become ever easier as it becomes easier to fake what an Atlantic story looks like. So we have the misinformation deceit risk, right? Someone will make a deep fake at an Atlantic staffer saying something crazy. Then you have a real risk. Something crazy could happen. An AI can facilitate all the terrible things that can happen. And so I won't go into those specific risks. So I do think about them a lot. But in any case, it's harder to maintain reputation. But if you can maintain reputation, you're in a better spot.

39:41At the same time, there's going to be like creators are going to have all these booths. It's going to be so much easier to be a creator. It's so much easier to be a good writer now, right? It's so much easier to get a good edit. It's so much easier to find stories. It is for everybody, so it makes it more competitive. But there are going to be many more creators out there. And then, as you say, no one will know what is real. I fail all those New York Times quizzes. Do you take those ever? It's like, what's an AI-generated image and what's real? I regularly get like 40 % results or 50%. It's impossible.

40:07It's impossible. And so I guess the most interesting implication of this is like, does social media die? Is it fun to go to TikTok when you have no idea if it's real or not and if it's a person or not? Maybe it's more fun. I kind of think it might die. I kind of feel like Twitter, TikTok, social media might just be destroyed. Like it'll be like you're in Westworld. And 10 years ago, you see someone talking, you know it's a person. Two years ago, well, you know, you won't know it's a person. You won't know if they said that. Even if you see it's a person you recognize, you won't know if it's them.

40:37You know, I have a friend who does daily videos like my most interesting thing in text. And once a week, it's him. And four days a week, it's his avatar who looks just like him. And saves him time. Yikes, right? And so the question will be, if social media dies because it's Westworld, where will real humans go? Will they go to WhatsApp chats? Will they go to forums where you have to actually pass key identification? Will we just gather in person? So I don't know exactly what happens. If it's entertaining, though, do you think people care? I don't know if you followed this, but Love Island did like someone made an AI version of it with fruit.

41:10Did you follow this? And it went like super viral. People loved it. So like, is there also an argument where people like might not care? And if they're entertaining and it's interesting information, maybe they do stay on TikTok. Maybe. Maybe. Will they try to understand the world? I feel like there's a real need for like human connection and for understanding real information about the world. So maybe there's a place where TikTok becomes like infinitely entertaining because like. It's just slop. The algorithm looks at your retina, sees exactly what you want, and it knows so much about you. It's like, you know, they want to look at this kind of person doing this kind of thing.

41:46They want a humorous video with a cat in it and a hot person in a bikini. And so then suddenly that image is made for you. And this person wants an economic minister from a European country denouncing a social media company. We'll make that for them. Maybe that's what TikTok becomes. But then there are other people who are like, you know what? I just, God, help me. I just want to talk to real people, right? And so then they have to go somewhere else. And that's where we get back to the events business. Getting back in person. Events business and in person and also group chats. And also physical.

42:16I think there is something about having the physical magazine. There's so many creators. It's so funny. They get like a New York Times cover print edition and they like show a picture of it because it's something tangible about your success. That's why you see, you know, YouTubers going for Emmys now. It still matters. These awards, the recognition, the physical, it still matters even in the digital world, I think. Yeah. It was actually cool to see Jorge Ramos, right? He went from winning all the Emmys at this TV station to winning an Emmy as a creator. Yeah. Thank you so much, Nick.

43:03Thank you.

43:36Thank you.

44:08Thank you.

44:42Thank you.

45:01Thank you.

45:31platform matters. Automated recovery, clean data, a business that keeps moving no matter what hits. One platform, not a patchwork of stitched together tools and gaps. Don't wait for the next attack. Secure and accelerate your business at rubric.com. Again, rubric.com.