Ben Smith on the Semafor World Economy Summit: AI anxiety, labor market self-censorship, and the disconnect between private and public sentiment

Speaker 2: Well, we have the beginning of our guest lineup starting now. We have Ben Smith from Semaphore. He is the editor in chief. He's in the waiting room. Let's bring him in to the TBPN Ultradigm. Ben, how are you doing?

Speaker 1: What's going on?

Speaker 9: Doing great. Can you guys hear me alright?

Speaker 2: We can hear you.

Speaker 8: Clear.

Speaker 2: We can see you, but we

Speaker 1: don't know where you are.

Speaker 2: You're looking very sharp.

Speaker 1: Dapper, I might say.

Speaker 9: Thank you. That's you know, we're convening a biggest gathering of business leaders in The US this year here at Summit Four World Economy. Wow. And so How

Speaker 2: many business leaders is Yeah. How many people are attending?

Speaker 9: 500 global CEOs and then

Speaker 1: Wow.

Speaker 9: Lots of political figures from The US and Europe. Had Scott Besant earlier, we got a Lutnick later. I assume they all just kind of came over from Hill and Valley. Yeah. We do our best.

Speaker 2: That's great. What's the structure of the conference? It's a few days. Are there specific breakout sessions, talk tracks, certain themes, or is it sort of more driven by the individual who's on stage at the moment?

Speaker 9: You know, it's five days, three consecutive three simultaneous stages, a bunch of really interesting kind of Chatham House rules, breakfasts and dinners and things like that. And I mean, the core topic is really, I mean, kind of where we live is the intersection of politics and and and sort of political power and finance and technology. So Sure. I mean, that's that that is what that is what people here are obsessed with.

Speaker 2: Yeah. And and within that intersection, what are the top topics that people are discussing in particular? Obviously, the the the conflict in The Middle East is big, but then AI is big. What what's actually driving the conversation today?

Speaker 9: You know, every conversation winds up coming back to AI. Really? And yeah. And I think whether that's, you know, Bessen actually, I had an interesting conversation with Bessen on stage where he he Yeah. Really downplayed the conflict with Anthropic. Okay. And so, they have kind of like some minor technical issue with the Department of War.

Speaker 2: Oh,

Speaker 9: interesting. The Treasury is working with them to make sure, you know, the banks don't go down. Sure. But just, you know, I think there's just a sense in particular that like really big enterprises and the kind of big brand name American companies are the ones most at risk

Speaker 10: Yeah.

Speaker 9: In the transition. And and I think people who work for those companies, who advise them, who compete with them are thinking a lot about that.

Speaker 2: Yeah. What about on the labor side of big business? We've seen some reasonable strength in the job market recently. At the same time, there's a lot of layoffs going on. How are people grappling with the with the the unemployment situation in the economy?

Speaker 9: Yeah. I mean, you know, the numbers are good. Yeah. And yet, there is this sense, you know, among the American public that AI basically exists to put everybody out of work. Yeah. And they hate it. Yeah. And so, that's lingering over things. And I do think there's also a sense that the class of 2026 is not a great time to get out of college.

Speaker 2: Yeah.

Speaker 9: And I think there is a I would say like both like kind of authentic nervousness about, you know, what is everybody's labor force going to look like? Because of course, class twenty twenty six are also the ones who are native to these tools. Yeah. And are the ones you ought to be hiring. Yeah. And then also of like, they gonna come and chop everybody's heads off when they realize they aren't getting jobs? So I would say those two two competing questions about the youth.

Speaker 2: Has there been any discussion around resolutions to that or economic policy that can strengthen the safety net or or, you know, deal with taxes or more weekends or anything that can sort of offset the anxiety about weakening labor markets?

Speaker 9: You know, every member of Congress is, like, thinking about this all the time, taking questions about it all the time, and yet also aware that this is an institution that spent twenty years talking about regulating social media, and only now is sort of getting to it, but most of it's happening in The States. Yeah. And so, I think there's a lot of pessimism that Washington will act. But I think there's also is a sense that I think a lot of companies are seeing unexpected jobs coming out of AI. Talked to somebody who operates, like, basically the sort of call center dealing with financial fraud that you would think would really get wiped out by AI.

Speaker 2: Yeah. Totally.

Speaker 9: But the good news is there is so much more financial fraud being driven by AI that they're keeping people very, very busy.

Speaker 1: That's emerging in in the legal space is like you just get way more contracts because contracts are easier to create or you get more lawsuits. So a lawyer is spending less time per lawsuit, but there's a lot more of them to work on. Yeah. And so you still have Yeah. That's full

Speaker 11: The good news.

Speaker 2: Yeah. How are you using AI in the newsroom at semaphore? Are there I imagine you're not just writing AI articles, but are is there internal tooling that's been built? Are you understanding

Speaker 1: that metrics better? Are well set up for this because you're there's like, hey, here here are the facts then here's like takes and like separating those two things out. Yeah. I don't care what you use here. I want I want the takes to be generally pretty organic and coming from the author. But how do you think about it?

Speaker 9: Yeah. You know, we use it and it's interesting because I think like a lot of media, there's these two sort of huge trends. One is AI and automation, and the other is creators and individual connection. And people do not want AI in the way of that. I mean, when you when they realize that you're just avatars, you're gonna be in trouble.

Speaker 2: For sure.

Speaker 9: And and so we use it all all over the place behind the scenes to try to like help amplify our journalists. I I vibe coded something that predicts mean tweets about stories that I'm personally very proud of though I cannot say that I have driven wide adoption with my with my team. Yeah.

Speaker 1: Get pre dunked on, use this tool.

Speaker 9: That is what the tool I called it pre dunk.

Speaker 2: Pre dunked. No way.

Speaker 9: They spent a whole day. It's in Google Apps Script. I love it.

Speaker 2: That's awesome.

Speaker 9: I would say more more usefully, one of my colleagues, because actually one of the challenges with an event like this is you've got three simultaneous stages and the reporter who's most qualified to write the story or even to know what's news is often the same one doing the interview and sitting on stage.

Speaker 2: Sure, sure.

Speaker 9: And so we did build a great tool that, you know, just takes the transcript of every interview instantly and pops out what's new. Oh, First so that a reporter who's maybe more of a generalist

Speaker 2: Totally.

Speaker 9: Will know, oh, yeah. This is news. This isn't, which is incredibly useful. And we, know, obviously we check it, but that's a great starting point.

Speaker 2: Yeah. Yeah. That makes a lot of sense because there's so many times when people fall back to a talking point, but you're looking for something that's actually novel or a new data point or a new prediction or something announcement. And that just might sneak in.

Speaker 9: Get this. Like, do you

Speaker 12: deal with it? You do

Speaker 9: an interview and you think there's one part that's interesting, but you're on the air and something goes viral.

Speaker 2: Right? Totally. Totally. Yeah. And we've and we I that's a very interesting tool. We should probably think about integrating something like that because we yeah. We found that it's very, hard to predict even with the best AI tools what content, what clips will actually perform, what is news. And we see a lot of coverage downstream of our show where something that's set on our show will then turn into coverage on other news sites, and, we've never really figured out how to grapple with all of that, but that's that's interesting.

Speaker 9: It's not just performance. Right? Like, it's also you might, like, think this is the new part. This is the interesting part. Yeah. But if you're doing it live, it just sort of floats by. Yeah. So that's been that's probably the thing that I'm just this week that I'm most kind of jazzed Yeah.

Speaker 1: What about overall market sentiment? J.

Speaker 9: D. Capaluno who made it. Sorry.

Speaker 1: What about overall market sentiment? I don't think if you told me last year we would be, you know, over a month into a war with Iran and the S and P five hundred would be at a record high. Yeah. I would not have believed you.

Speaker 9: Yeah. I mean, it's amazing. And, you know, people who CEOs, people who show up at events like this tend to be optimistic people. It's sort of a job quality, and so there is, like, a pretty positive vibe here and being driven there sort of parallels and is driven by the markets. One thing about Washington now, though, is there's also a lot of self censorship.

Speaker 2: Mhmm.

Speaker 9: And there's a real difference between what people say in public and what they say in private. And I think, you know, in our in our kind of Chatham House rule gatherings, you're hearing a lot more concern about a kind of about a bit just about the long term direction of of US power, The US economy, The US markets, people hedging away, and no one's saying that stuff in public.

Speaker 2: Yeah. Well

Speaker 1: Yeah. Biggest biggest ever disconnect in my life between like group chat conversations and like timeline conversations among people Yeah. Who have something to lose is how I would put it.

Speaker 2: Well Yeah. Interesting times. Hopefully, it, what's the plan to turn the synthesis of what you're hearing in the Chatham House context into reporting or content? Will you be writing op eds that are informed by what you learned? How how do you think about actually distributing that information?

Speaker 9: Mean, basic Chatham House rule is that you can steal the ideas and pass them off as your own, but you're not allowed to contribute them. So Yep. That's the plan.

Speaker 2: Okay. Well, we'll look forward to it. I'm sure everyone can follow along at semaphore. Thank you so much for coming on the show.

Speaker 1: And Yeah. We'll make it next year.