In short
Big Technology Podcast Episode Notes
Episode Title Can The News Industry Survive The AI Era? *Featuring Ben Smith, Nayeema Raza, and Joe Marchese*
Episode Overview In this episode of the Big Technology Podcast, host Alex Kantrowitz engages in a dynamic discussion with Ben Smith and Nayeema Raza, the hosts of the new media podcast Mixed Signals, and Joe Marchese, a partner at Human Ventures. The conversation revolves around the challenges and prospects of the news industry in the age of generative AI, particularly in light of recent multimillion-dollar content deals with companies like OpenAI. The episode draws parallels with past media technology partnerships, critically examining whether such arrangements can lead to a sustainable future or repeat past mistakes.
Key Discussion Points
Historical Context
- Previous Partnerships: The discussion begins with a reference to past partnerships between media companies and tech platforms like Facebook and Google, highlighting how these deals did not lead to long-term stability for journalism.
- Comparison with Current AI Deals: Guests express skepticism about current arrangements with AI companies, suggesting they may similarly fail to provide meaningful value to news outlets.
Concerns About AI's Impact on News
- Danger of AI Reliance: Concerns were raised that partnerships wherein news agencies receive compensation in credits (a non-cash form of trade) may not foster true innovation or sustainability in journalism.
- Monetization Issues: The guests discuss the fundamental problem of monetizing news effectively in an AI-driven landscape, where AI can provide quick answers that diminish the value of in-depth reporting.
Future of News and AI Collaboration
- Need for a New Business Model: A suggestion is made that the news industry may need to adopt a consortium-like model to negotiate better terms with AI companies, akin to how Spotify operates within the music industry.
- Potential for Individual Brands: The conversation touches on the rising trend in journalism where individuals are building their brands, potentially creating a space for personalized news experiences in contrast to AI-driven generic news.
Ethical and Legal Implications
- Copyright and Intellectual Property: The discussion brings attention to the legal implications of AI using news content to train its models and the potential for lawsuits over copyright issues.
- Public Trust in Media: The importance of maintaining trust in media brands is emphasized, suggesting that as AI becomes more prevalent, there could be a market for trusted news sources that offer verified information over generic AI-generated content.
Consumer Behavior and Media Consumption
- Changing Dynamics: The conversation reflects on how consumer behavior is shifting as audiences become increasingly wary of traditional news sources and seek out more reliable, personalized content.
- Nostalgia for Traditional News: Acknowledgment of a desire for a return to a centralized source of news that everyone can engage with, similar to past eras of journalism.
The Future Landscape of AI and News
- Competing Perspectives: The potential for a hybrid environment where users have a choice between AI-generated summaries and traditional journalism, raising questions about the viability of both models.
- Impact on Journalistic Integrity: Concerns are raised regarding whether AI can effectively replace the nuanced, investigative work that professional journalists provide.
Key Takeaways
- The news industry faces significant challenges in adapting to the AI era, with skepticism about whether current partnerships will yield positive results.
- There is potential for a new collaborative model between news organizations and AI companies, but trust and ethical considerations must remain paramount.
- Individual brands and personalized news experiences may thrive, providing an alternative to generic AI-generated content.
- The conversation hints at a future where consumers may value trust and brand integrity more than ever, creating opportunities for media outlets that can navigate the changing landscape.
Conclusion This episode of the Big Technology Podcast offers a deep dive into the intersection of AI technology and the news industry, exploring the complexities and nuances of how both can coexist and potentially thrive. The insights from industry insiders encourage listeners to consider the future of news consumption in an evolving digital landscape.
---
Additional Resources
- Podcast Link: [Big Technology Podcast](https://www.bigtechnologypodcast.com)
- Mixed Signals Podcast: Available on various podcast platforms.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00Let's talk about whether the news industry can survive the AI era. with a venture capitalist and three journalists. Oh, this is going to be fun. All that and more is coming up right after the break.
0:13You're used to hearing my voice on the world bringing you interviews from around the globe. And you hear me reporting environment and climate news. I'm Carolyn Buehler. And I'm Marco Werman. We're now with you hosting The World Together. More global journalism with a fresh new sound. Listen to The World on your local public radio station and wherever you find your podcasts.
0:40Welcome to Big Technology Podcast, a show for cool-headed, nuanced conversation of the tech world and beyond. We have a great show for you today. We are going to talk about AI and news and also what's going on with the media industry because there's a lot of big stories there and it's going to be covered in a new podcast by two of our guests that started on Friday. Ben Smith is here. He's the editor-in-chief of Semphor and he's the co-host of Mixed Signals. Ben, welcome back to the show. Always nice to see you, Alex. Likewise. And we're also joined by Naima Raza, who is Ben's co-host on Mixed Signals.
1:14Naima, great to meet you and welcome to the show. Thank you for having us. And we also have Joe Marchese here. He's a partner at Human Ventures, a VC that will actually speak with journalists. Look at that. Joe, welcome back to the show. Hey, how are you? We're speaking, this is Wednesday ahead of the show, and we're speaking on a day when both the Atlantic and Vox Media Inc. deals with OpenAI, effectively giving them their content and they can train on it and there's some credit swap potentially. And, you know, this feels a lot like the BuzzFeed days when we inked deals with Facebook and Twitter and thought that the news partnering with tech was going to usher in this new era of productivity and financial stability for journalism.
1:57And that didn't exactly work. Do you think that there's a chance this time will be different? Or is it the same problem just being repeated with the new technology? You know, I think everything is, you know, it's always different. I got to say, whenever you're getting paid in credits, that should make you a little nervous. It's not a good deal. I don't know if they're counting that in the numbers that they're releasing. Because what's attractive about these deals to the companies is that they're 100 % margin revenue. You know, they're just, these are all companies that either are losing money or kind of like really working to break even.
2:34And this is free money. but I do think that it's there's a real danger that comes with that I don't think it's exactly like I mean I think it's like Facebook in that a lot of the Facebook deals were intended to kind of politically mollify an industry that could do them harm and to buy off the journalism which to some degree worked to some degree didn't and what this is trying to forestall lawsuits but I think the risk and this is something our mutual friend Jessica Lesson wrote about the other day is that if you're sort of entering into an arrangement whose basic kind of like business value exchange is that we pay you and you don't sue us, you aren't going to build a real sustainable business together.
3:17Like that's it. It's a different, and it's too soon in the history of AI to even know what that news business would be. And so in a way, it's sort of foreclosing the idea that there's some valuable business to build, and instead there's just kind of like some quick extortion. Right, and this is Jessica speaking, writing in The Atlantic, I have it written down here, that media companies would rush to do these deals after being so burned by their tech deals of the past is extraordinarily distressing. It's hard to see how any AI product built by a tech company would create meaningful new distribution and revenue for news.
3:50These companies are using AI to disrupt internet search, to help users find a single answer faster than browsing a few links. So why would anybody want to read a bunch of news articles when an AI could give them the answer? Maybe with a tiny footnote crediting the publisher that no one that no user will ever click on joe you're coming in from the tech perspective so why would a tech why would a news publisher want to sign this deal with open ai well i'm somewhere in between and i do think this is lucy with the football uh with charlie brown that meme is going to live forever on this which is a platform saying they're going to pay news publishers and they're going to help fund the content but before they were helping fund content that then would drive more engagement on their platform and then maybe drive some traffic back to the publisher.
4:37Now it's free money, but then if we never send anyone to you, you can't complain about it because we'll just give them the answer on our platform in a business model that is unknown. So, I mean, it is incredibly speculative and I have no idea. I haven't gotten into what's inside these deals. How long are they for? Can they train? Because what's going to be most value is new news. Like training on the historical news, that's, I mean, horses left the barn. That's what you're not suing me for. But training on fresh news will be all the value going forward. And I do wonder what those deals look like.
5:07No, they don't look particularly good. I mean, there's a$250 million deal between News Corp and OpenAI. That's$50 million a year. The revenue of News Corp is$15 billion or so, right? So it's not nothing, but it's not a huge amount of value. And I think the big question is, you know, how long, what is the value of that training data? You know, are we in the, are we in the, where drivers taking inflated fees for subsidized rides version of Uber only to be replaced by, you know, self-driving vehicles in the future? Or are we somewhere else? But I think that's the difference between the training data, like historic versus news, like, you know, like before what the platforms did on their own.
5:51And by the way, not just Facebook and Twitter, but search. It atomized news so everything was next to each other. So it took a news story and then put it next to everybody else's news story. And the engagement occurred on their platform and they could monetize. Now what they're saying is, okay, trained to talk like news is one thing. Actually having fresh news is something different. And part of my thinking when I heard that was, why doesn't one of these platforms just buy the news organizations at this point? Like gathering fresh data from the real world is what they're going to need. Or more than that, why don't they just make a deal with one company?
6:26Like couldn't the AP, for instance, suffice? Like it does really seem like making a deal with all these different companies is basically like, please don't sue us. Whereas like even the News Corp deal alone probably could satisfy all of ChatGPT needs, ChatGPT's needs and more. It would make the system be vulnerable to what has been the greatest critique of kind of media in the last, you know, as long as I've been operating. It's just like bias, right? Having a single source. So having, there's a value in having multiple sources. It's also a value in kind of understanding how things are done across different media outlets and different formats.
7:05I don't know. Ben, what do you think? Yeah, I mean, I do think, I mean, kind of what Joe is getting at is that I think that we had come in sort of with these metaphors out of social media, which among other, or these analogies, which among others, there are going to be one or two big winners and that they're really special. And then that news is a commodity and we're interchangeable. But I'm not sure that's how this landscape shapes out. Like it looks increasingly like these models are kind of interchangeable and there are going to be a number that are good enough and having exclusive facts might be really valuable.
7:35And I do think there's this assumption that's been around for 20 years on the internet, or at least since the social media days, that most of the revenue, most of the value go to the distributor. That YouTube is a big company, that Google's a big company, that Facebook's a big company, and that content creators can kind of eke out pennies and little percentages around the edges. But that's not really particularly consistent in the history of media. There are Other models, Spotify is the obvious one, where you build a really cool tech platform and your reward is a tiny, tiny margin and maybe zero.
8:10And I think if you're in the content industries, that's the world that you want. They can build their NITO distribution system and they should have the right to make fractions of a cent on every dollar. Once they've trained on the historic data, tone of news and analysis, that's how the models train. if you're going to have the value accrue, you need some sort of consortium amongst news to do that. Like Spotify is the model that would be better for the news world, but it doesn't even have to be an oligopoly. It would just have to be a consortium of agreement at the news layer. And that was what people talked about when search was coming around.
8:47And, you know, Rupert was saying, hey, we shouldn't be letting Google crawl this for free. And then Google won, obviously, by bifurcating everything. And remember the arguments back then. And the arguments were, well, hey, why is good content have to come from premium publishers? Isn't this person just as good? And we know that it's not just as good. So the question of, like, will that happen this time or will they just pick off enough and then say, oh, well, the open web will provide the rest. Can publishers not even do that? Like, aren't they legally not allowed to band together? No, absolutely.
9:21Like, no, consortiums are allowed in tons of industries. And there was one that was tried, actually, at Semaphore's launch event. Barry Diller spoke about this idea. And then publishers got together and tried to band together to some extent. They've fallen away, right? And Barry Diller's organization was one of the first to ink a deal. And now it's like the New York Times and some of these other organizations that are still involved and broiled in lawsuits. Now, the Times has kind of made an argument for billions of dollars of statutory damages for copyright, which makes the$250 million deal with News Corp look tiny.
9:58When Sam's going to raise$7 trillion, I mean, what's a couple billion? But the other point what Ben's getting at is if the distribution world is going to be more competitive than we believe it to be, and than it was in the kind of social media era, then in fact as long as these news companies aren't signing themselves into exclusivity, It makes sense why OpenAI would get into this business, want to pay a premium to get in fast, train faster, be first, and have a relationship. But these news companies can then go ink the same deal with Grok if Elon were interested or others, as the case were. Elon's not going to do any of those deals.
10:38That's for sure. I did think that point was interesting. I mean, Alex recently had a story where you had used Grok.ai, I guess reached out to Elon, and he'd gotten back to you. What I thought was so interesting about that was the citations seem to be, there is value in the discovery, and there's also value, I think, in the citations in a world in which trust has been really diminished and critiqued. And especially with this crowd, right, the early adopter crowd for kind of AI search, and this crowd which has been resistant to media in a big way. So citations have a very like are almost advertising for the core news product.
11:18Well, but listen, I did I did reach out to Elon about the citation thing. He said they're going to add citations to Grok. But this is very interesting, the way that they're training their Grok news summaries on Twitter. And I guess elsewhere when they start to develop Grok more broadly is that they're not even going to look at the stories. They're just going to look at the social media chatter and build their summaries from that. So they're not, they're removed from the actual legal liability of training on news stories because they're just simply talking about what people are talking about on social media.
11:48And to get a citation, presumably, and I don't know this, but I'm assuming he would need to post on X to get a citation in Grok. Is that correct? I mean, even if you're not posting on X, like somebody else is going to write about your story. It's out of your hands. That's a citation. And so it is interesting. And one of the things that I think we should all think about is like, does the economic system break down? Let's say this does become the way that people interact with news. It sort of becomes like, all right, you go to ChatGPT. What's happening today? It tells you all the things. Does it then remove the incentive for publishers to actually, you know, go out and report the news, which is expensive?
12:28Because you're ultimately like being disintermediated here. You're being presented as a summary. There's a citation. Let's be honest. Nobody clicks that. Like, is this what what the AI tech is going to do to the news industry is just continue to remove the value from actually people going to the sites and ultimately disincentivize the creation, making the AI products worse? You guys have all seen the Google generated results where there's a certain amount of rocks you can eat and there's a certain amount of non-toxic glow. I know. Well, this is this is the Simpsons, the Simpsons Guide to Diet.
13:02it. But joking aside, actually, this is my hope, is that then trust matters again. And I don't want to go to the generalized ChatGPT or Gemini for news, but I will go to the Semaphore AI news app, right? And because there's some object permanence to the brand that I'm trusting, and if that brand lets me down, then I'll move on to another brand. Whereas if Google lets me down, I can't move on to another search. Or if Amazon lets me down, I'm not moving on to another e-commerce platform. So the things that matter to us, actually brand could matter again. And that I do think is the great hope for AI and this.
13:39And that is where Ben's point would come in, where the value could accrue to the people who maintain trust at the consumer interface level, the consumer of news interface. Ben, is that even feasible? Like does Semaphore have the budget to go build its own AI that's going to compete with ChatGPT? Or do you really think that if Chachi Piti can summarize, let's say, your stories in the media newsletter or even your podcast, that you're going to get a good deal because you have the trust that, let's say, the Eat Rocks diet doesn't. I'm hoping that Joe has a fund with a couple trillion dollars in it for our compute.
14:15I'm hoping that too. No, but I think that in a way, in a world full of hallucinating AIs and malicious or just goofy deep fakes everywhere. I mean, there's obviously a question. And some consumers who kind of aren't looking for trust and view it all as entertainment don't care and are perfectly happy on X because it's all kind of a fun sport. But others who want to know what's trusted, I do think there's some return to a trusted brand who can tell you if that picture of Rihanna at the Met Gala is real or fake. There is some perhaps somewhat modest space for that. But it is, though, I mean, all of these things, I think, go to Jessica's point, though, of how little of this has shaken out, how totally unclear it is.
15:02I mean, among other things, you know, are these, is OpenAI going to be like Facebook or Google, which is to say, at least for sort of a moment, a kind of quasi-monopoly printing money? or is it going to be like Uber, which is to say like a company whose revenue can never quite catch up with what it has to invest because it's ultimately in a business with no moat that's really competitive. Can I take the Uber position here? I mean, one of the things that stuck out to me in the release that at least I read from Vox talking about its partnership with OpenAI. So you know both companies' eyes have been on this release.
15:37They said ChatGPT had 100 million users. I mean, ChatGPT had 100 million users two months after launch in December 2022. You know, by my watch right now, it's May 2024. And I think in these conversations, there's often a presumption that this is just the way things are going to happen, right? That people are going to want to read news through ChatGPT and that AI overviews from Google mean the end of the web. But the consumer behavior just hasn't caught up to this vision yet. Even Bing, which has been serving AI answers in search for how long? Like more than a year? has 3.6 % of the search overall share in the world.
16:14So this was actually so much better than actually like listening to shows like this one or make signals or reading stories like the ones on Semaphore or on big technology. Like maybe we would already - Yeah, I mean, we got to shout out where we're up against a pretty strong force, right? This is human level intelligence we're talking about or close to it. I mean, that sentence, just that sentence, even Bing and then being like not big enough. It was a sentence I would not think I would have heard in 2024. But we're past the like Bing sucks because Bing is good now. Like it's actually good. And it has GPT for all.
16:47Do you think Google is making a mistake by sort of racing into this world that like the press and the sort of AI visionaries have decided is the future? This sort of post link chat bot for front world. I mean, I think that Google claims that they're following consumer behavior. I mean, do you, I don't know. do you think google even has free choice in the matter because if they weren't doing this what would the market be saying about them like ever since they started to say hey we're all about ai they've hit all-time highs in the market whereas beforehand you know they were they were getting pummeled yeah irrespective of the quality of their product the stock price is going up yeah and and i mean i guess your point alex was kind of like they have no choice because everyone else is going to do it and maybe more people would be searching on bing if google that doesn't have that.
17:34But I do think that the question of, like, our news has been being served to us by AI for a long time. The TikTok algorithm, the search news feed, the Facebook news feed, like AI has been picking our news stories for a while and kind of famously, that's going to come to a head in the courts pretty soon. But now it's just that AI is synthesizing news. And I think when consumers don't have to pick through and they actually have to make a selection of what brand's AI news do they want, is an opportunity to seize on. I don't want to ask ChatGPT what men's fall fashion are, but I would ask GQ's AI what men's fall fashion are because I know that they'll have inputs from editors and experts.
18:10Or you might ask ChatGPT what GQ suggests, right, if that's the interface. You've asked the question better. I mean, I only wear black T-shirts, so it's going to be really hard for me to figure out which ones to get next season. But I will say that the, and my hairstyle is not going to really need much, but I will say that the idea of if you're going to use a brand name in your search, That will be an ongoing revenue stream, not a training on past history stream. That's really interesting. Yeah, the currency matters more, like the recency of the information. But then also, can you imagine a world in which AI would be doing reporting?
18:47That's a good question for Ben. I mean, I think there are certain kinds of fact-gathering. You know, there are times when I, like I used to, you know, like there are times when you are trying to pretend that you are sending customized emails and you're in fact copying and pasting them into 400 windows and hoping somebody writes back to you. I do think there are like labor saving hacks. And, you know, and actually my colleague Gina Chua, who's done a lot of pretty ambitious playing around with this, basically there's no national database of hate crimes. and what each jurisdiction says is a hate crime is kind of different and articles describe them in complicated ways and she basically had chat GPT or OpenAI go through a huge corpus of news articles and say how many hate crimes were there in this period and it did a pretty good job this stuff that is not exactly black and white but doesn't require that much analysis and involves reading huge amounts I mean, I would not then trust that and say that's something I can publish.
19:53But I think there are kinds of synthesis and of kind of like reporting on just like large amounts of information that it could be useful for. I think the, you know, the core, the thing of getting somebody to tell you something that they weren't supposed to tell you is probably like that's probably still, that's defensible. One of the biggest trends we've seen in journalism to date in the last 10, 15 years has been the establishment of individual brands. The fact that I trust something because it's by Ben Smith. I trust something because it's by Alex, whatever it is. That, to me, is where AI synthesizing the news, depersonalizing it.
20:39And consumers are frustrated by the reductiveness, the lack of nuance. So I actually feel that the idea that this is going to be like the bringing back brand, I don't know. It's almost like you would ask very specific questions. What does Maggie Haberman say about X because I trust that or I don't or whatever it is, right? Yeah, those are the trends in opposition. There are two trends sort of happening at the same time, right? One is the creation of these anonymous giant synthetic models, and the other is this huge flight toward individual brands. And they're obviously kind of going in different directions, although maybe the place they land is that you have one superstar blogger, Alex Kantrowitz, who can use AI tools to produce his television show at pretty low cost and not employ anybody.
21:26And yet seem like he's running a giant media company, which is just him. But yeah, the institutions are the ones that are inking the deals, right? And they're the ones up against both of these trends. The individual journalists are not up. Right, and the Wall Street Journal is the thing that's getting squeezed there. Yes, exactly. But for now, like why isn't there talent agents or IP deals for individuals? Because, you know, I'm guessing Casey Newton, a platformer, people like, okay, I like your voice, like the way you cover these things. I want it in that voice. I do think that there's a place where the individual brand comes in and I pay.
22:02It's just the same way talent comes in every other industry. Yeah, but a lot of those single-use apps failed in the YouTube era, right? Because in the YouTube era, there were companies like Download and every YouTube influencer under the sun from Kim Kardashian to Alden Rose was going to have their own app, and very few of those succeeded. That is because those platforms, look, this is why I started, the platforms are like Lucy with the football with news. It never is a better business model to pay for content than not pay for content, right? So it's what I'd call the mullet strategy, business in the front, party in the back.
22:35So I'll pay for this, but really I make all my margin of zero cost content on the back end. Well, there is no zero cost content on the back end if the consumer is choosing what voice they want to listen to. So every time you went to YouTube, I might come in for a New York Times article, but I end up on 14 UGC ones where YouTube keeps 100%. That doesn't happen in AI. And I can't tell you I know what the business implications are because there is no business modeling on their side yet. But that's at least an opportunity. I'm thinking often about this Scarlett Johansson thing where she had this feud with OpenAI.
23:05They wanted her voice. She said no. They used her voice anyway, but kind of slightly different. But from a different actress who was like, be Scarlett Johansson. The actress was like, sure. And here. And then Sam Altman tweets her. And Scarlett Johansson's like. Like, yeah. But this is the thing. We talk about, like, we have these discussions. Like, can the media survive this AI era? Can news survive this AI era? Is AI going to, you know, do creators have even a semblance of power against what's coming here? And it just seems to me like, I'm curious what you think, Joe. Doesn't the user always win here?
23:42Like, the user will get what the user wants. If the user wants Scarlett Johansson's voice, the technology just will progress without very much of the guardrails hardrails or even consideration about what happens. It might take a couple of leaps to get there. Like opening, I would win a lawsuit against Scarlett Johansson without a doubt, right? They paused it because they didn't want the PR headache. So do we, is there, is there actually any slowing this stuff? And I think about it also when it comes to this, like alignment stuff and super alignment stuff, like, yeah, it's nice to have a super alignment team within open AI or Anthropic, but like, are they actually going to do anything to slow the progress?
24:17That attitude is only very recent. Before Section 230, there was libel laws and infringement, and there was all sorts of issues with doing something like just taking Scarlett Johansson's voice. And so this is an attitude that is just of recency for the platforms, because the platforms throw their hands up and say, well, we're just a platform. We can't help it if those other people are frotting Scarlett Johansson's voice. AI is generating the response. So it's no longer like the safe harbor was for hosting content that wasn't yours. You didn't have rights to once you're generating it. I don't think that that's the same thing.
Read the full transcript
24:56And, and the reason why you say they would blow through is because we haven't had any, there's been no courts and legislation that has done anything about that for the past, I don't know, decade, 20 years, maybe. And, and so I think that this is a different time. And like the opportunity is huge, like for this to actually be good for copyright law and for trademark law and for personal image rights. And maybe this does that. But they used a different actress. Like how do they end up on the receiving end of a lawsuit and losing? Well, let's see, what they claim is that they use a different or they have said is that they use a different actress even predating.
25:32So imagine they created some kind of beta or trial version and then just went to market with something like that, that presumably was attracted to them because it sounded like Scarlett Johansson. Now, legally, Scarlett Johansson can still sue. I'm not on team copyright theft, by the way. I'm just like looking at the facts and asking. No, I'm just asking. I'm just asking questions here, people. I know. You are. You are. But Scarlett Johansson still can sue under the publicity, right? Under whatever it is, copyright publicity. Well, the her tweet was pretty ill-advised. Exactly. And it doesn't look good.
26:03But I think you said the answer in the first, which is they've lost in a court of public opinion. And consumers are going to care. I think Ben has made this point. AI is closer to the culture industries than most technologies have been. And the way in which we've seen people win against them are either single big brands like Scarlett Johansson, you know, threatening a lawsuit, but also really using her own voice, her own power, to kind of shift the spotlight of the conversation, the regulators, et cetera, to pay attention to this issue. And the other side is in the strikes that we've seen, WGA and particularly the SAG strike around AI, it requires massive organization and negotiation, which is probably easier done right now and easier conceived right now given the technology is nascent and there's one or two or three big players in the space versus when it becomes, if it becomes, Alex, the next wave of the internet.
27:00at. Is there a way here where everybody wins? I mean, the New York Times has built like this. I'm curious if anybody has some thoughts about like what's going on there, but they've built this big AI unit that's being headed by Zach Seward, who's a very well experienced industry insider who knows what he's doing around tech. He has a full team there. They're trying to harness AI for good. Like, is there a way the news industry wins here, which is a question we don't really ask often, but I'm curious what you all think. I mean, there's a world where the New York Times wins, I think that they're the ones who have benefited from the sort of centralization of these internet businesses.
27:36And they're the Google of original news in a way. And so they're in this dominant position, which they can use to become more dominant. Yes, definitely. Also, in other ways, a very talk short AI. They produce a lot of words. They have a huge body of text. The companies really need them. like an AI that was missing the New York Times would be lame and would be worse than the one that had it and they kind of know that and the Times is going to hold them up for huge amounts of money I mean I think the question of like you know I mean I think some of this stuff is sort of a little orthogonal to the news business like a lot of news is your local community weekly which was getting killed in the social media and the sort of Google era and will now get killed by this but needs a different model I think AI is sort of closing the door on some of these ideas about just like building massively, rapidly scaled news businesses that we came up with 10 years ago and that we all kind of worked inside.
28:41And that sort of the kind of divorce with social media and the shifts in search had already kind of put to an end. But AI just means, I mean, assuming these AI platforms are sort of the next wave of platforms, very hard to see. I mean, I think in some ways it's going to help. It's going to entrench the advantages of big incumbents. I mean, the other people it shuts out, and Max Tani had a great story on this the other day, is partisan media. Like nobody's calling up the Daily Wire for a licensing deal. Nobody called up HuffPost, actually. Nobody called up Breitbart. And they're all grumbly about it.
29:16But what this means is if you go looking for, when you went looking for news on Facebook, you often got really, really insane stuff. Google, somewhat less. But when you go looking for news in Gemini, and it's just going to tell you the earth is round and Biden is president and stuff. You're not going to, like a lot of these people who thrive by saying the most extreme and exotic stuff will find places to do it. It looks like X will be increasingly that because it's so engagement-centric, but it's a smaller world. Does anybody think that the fact that we had news feeds, we had all this other delivery, and then email newsletters, like, come on, the boom of email as a delivery mechanism.
29:53And it was because the publisher controlled that connection to the reader, and the reader could know that that came from the person who published it, and it wasn't being interfered with anyway. And email is, what, 30 years old at this point? And it's still the preferred. So, like, what will that be for AI? Yeah, but email newsletters have also picked up, particularly in a time where there's been more news disengagement, right? Emails are coming to you, you're not going to them to some extent. You sign up for them, but then they continue to come to you. I do think that there is a - So did my Times paper on my doorstep, actually.
30:27Oh, I delivered them, but you know, same, same. I do think that there is something to that. There's a nostalgia about the news. People are a little nostalgic about how they get their news. They're a little distrustful. The platforms, the social media companies themselves are, you know, Meta would like to see less news on its platforms, which in fact is a good thing for a competitive advantage, arguably for open AI, et cetera. I do think what, you know, it will be curious what Elon does here because with X, I mean, the real value of Twitter is the breaking news. When the earthquake happened, there was nowhere to go but Twitter, right?
31:05Like no institution. I was on Twitter mid-earthquake being like, oh, shit, this is definitely an earthquake. Yeah. This is a great Twitter moment, right? As irrelevant as you think Twitter is, in that moment, that's the only way you're going to find the answer. So that combination of kind of breaking news and the tonality, the conversation, the zeitgeist that currently Twitter has, no other social media company has it right now. But you go there for hundreds of thousands of voices telling you what's happening in real time rather than one synthesized voice pulling it all together, right? Authority is lost.
31:40Yeah, exactly. It's homogenous and it's all mushed together and people's brains don't work that way. We need to tell ourselves stories. We need to trust where it's coming from. We need to get single person narratives. And that's what Twitter really provides. This is beautifully setting up like a question for the group that we talk about on the Friday show every now and again, which is what's going to be, what would be a healthier environment for news? Like news mediated through the screaming, you know, emotion-based social media feeds or news mediated by AI that is more even keel, but wrong fairly often.
32:16Or not fairly often, but wrong every now and again because it hallucinates. So would you rather your news be like super polarizing and insane but correct or sort of soothing and wrong? I mean, I think that's why you've got to read Semaphore where we are soothing and correct.
32:33That's amazing.
32:37The thing is, we don't get to decide. I mean, the thing with all this stuff is we don't get to decide. I think consumers are obviously really, really sick of hyperpolarizing, screamy, manipulative news, and you see it in the cable ratings, and you see it in Facebook getting out of the news business, and you see it in Twitter's sort of decline in marginalization. And not that there won't always be a market for that, right? There always has been, always will be. And maybe the pendulum will swing really hard back that way at some point, and there'll be another sort of populist wave. But right now, people are obviously retreating from that and looking for something different.
33:11That's sort of the business that we're in right now is trying to give a kind of more nuanced view with the kind of space for disagreement. I think a lot of podcasting right now, like the reason it's popular also, is because there's this kind of like non-hysterical, leaned back, open conversation. And again, that's not going to be universal, but I do think it's just sort of obviously where people are now. I think I agree with Ben on that. And I also think there's a wrong answer is with skepticism, because you have to bring in the consumer, is better than a diversity of opinion without any skepticism about your aunt or your boyfriend's opinion.
33:48So I think that the fact that people are going to look at this technology with some skepticism, even when it's serving a single answer, and then ideally go double-click, and again, that's where the value of these citations comes to be and where some of these news organizations might find value in discovery and citations, that could be a more productive world. But again, I think the most soothing is for them to go to semaphore with its very soothing kind of manila background. Can I give a third option to that, Alex, which is one I really want to see, which is just people know they've seen the same thing, whether it's right or wrong, right, or polarized or not polarized.
34:26I can't talk to somebody else if I don't know if we've all seen the same thing. And I do think that the misassumption right now is that, oh, I'll want personalized news, but then how do I talk to Alex about what happened today? And how do I talk to Nima? So what I would take is, the reason why we talk about Taylor Swift and Travis Kelsey and or the weather is because it's the only thing that's guaranteed that we've all experienced on any given day. And so the idea for news being like, everyone jokes about the Walter Cronkite days. Yes, there were problems with those news days. There were tons of top-down issues.
34:58But everyone saw the same thing, and it was like, okay, at least we can discuss that. So is there an offer in the making for, at least I know we've seen the same thing? I don't know. Yeah, that's the value of the front page. It's one of the pieces of information it gives you is what everybody else saw. Yes, right. Yeah, and then I do think you're impersonating, because in a way, I mean, another word for AI in a way is in this space is personalization. And it will be interesting to see how much of that people actually want. Okay, I need to go to break. Why don't we go to break and come back and talk a little bit more about mixed signals and pick up this idea about whether hyper-partisan media can survive.
35:34And is it going to be BuzzFeed when Vivek Ramaswamy runs the publication? All right. Why don't we do that when we come back right after this? Did you know your credit card points and miles can lose value to inflation? Credit card companies often reduce the redemption value of your points and miles. Now, imagine a credit card with rewards that can grow in value. With the Gemini credit card, you can earn Bitcoin or one of over 50 other cryptos instantly with no annual fee. Every swipe at the store or gas pump earns you instant rewards deposited straight to your account. Plus, sign up now for a$200 Bitcoin bonus to kickstart your rewards.
36:12Visit Gemini.com slash card today. Check out the link in the description for more information on rates. Again, if you're looking to invest in Bitcoin but don't know where to start, the Gemini credit card makes it easy. The Gemini credit card is issued by WebBank. In order to qualify for the$200 crypto intro bonus, you must spend$3 ,000 in your first 90 days. Some exclusions apply to instant rewards in which rewards are deposited when the transaction posts. This content is not investment advice and trading crypto involves risk. The Gemini credit card cannot be used to make gambling-related purchases.
36:49You're used to hearing my voice on the world bringing you interviews from around the globe. And you hear me reporting environment and climate news. I'm Carolyn Beeler. And I'm Marco Werman. We're now with you hosting The World Together. More global journalism with a fresh new sound. Listen to The World on your local public radio station and wherever you find your podcasts.
37:17and we're back here on big technology podcast what a great group we have here today we have naima raza she's the co-host of mixed signals ben smith is also here he's the editor-in-chief of semaphore also the co-host of mixed signals which is now live first episode was friday you can find it in your podcast app of choice and joe marchese is here he's a partner at human ventures Ben, let's just talk quickly about the Vivek news. So Vivek Ramoswamy seems like he's pulling an Elon Musk on BuzzFeed. He's acquired a high single-digit percentage of shares and is talking about the way the company should be run.
37:52What do you think is happening there? I mean, I think that much like his presidential campaign, he has sort of found a formula for getting a lot of attention and no votes. Ultimately, this is a controlled company. There's no formal way for him to exercise the power of those shares because Jonah Peretti, the CEO, still controls most of the voting shares. At the same time, his ideas aren't. That letter reads pretty well the first time you read it, and then you read harder, and you're like, wait, this is a pretty cursory reading. His core solution to media is to pivot to video, which is obviously something nobody has tried or thought of before.
38:33And he complains, and somebody went and Googled like complaints about BuzzFeed's news. And so he has like the world, there was an expose of the World Wildlife Foundation partnering with local militias in Africa who were murdering people. And so he's like, well, they were too mean to the World Wildlife Foundation, which by the way, his people, conservatives, totally were very into that story, which was true. And the organization expressed great sorrow, but he didn't click the link. It was a bunch of stuff like that. My conclusion here is that the only solution is for him to hire me to come back and be the editor-in-chief of Right Wing Buzzfeed and set things right.
39:15I know where all the anti-roke bodies are buried. Oh my God, please don't publish my drafts, Ben. Yeah, that's where they are. They're in your drafts. All those places where you would be like, I want to take down Elon Musk. He's so terrible because he's so anti-woke. And then I would go in and, you know, make sure everything was fair. Exactly. Because that's how New Zunes works. Thanks for doing that. Yeah, hopefully they don't get those Google Docs. Those millennial journalists trying to take out upstanding reactionaries. So the podcast is interesting. I mean, we do need, there is room for a podcast about media.
39:51And Ben, you like in the announcement. I'm going to put that on like the label. There is room dot, dot, dot for dot, dot, dot, a podcast. Alex, can't do it. What happened to the good ads in the first half of this episode? We actually discussed that. Oh, my God. Wow. You can hear more discussion about Ben. This is the part where we just roast you, Alex. Sorry. I'm just setting you up to talk about what you're trying to do in your podcast. No, no, please do. Thanks. Naima, take it away. You're the nice one. Well, no, I was going to say we have a conversation. We get a little bit deeper into Ben's portfolio and resume for his BuzzFeed job lusting in the episode that you can hear on Friday.
40:28Friday's episode, we talked about it at great length and something spicy there actually to check out. But there is room for it. I think that people, it goes back to brand. It goes back to people just looking at this news environment. And we're seeing simultaneously people think this is the most important election of all time. I'm just giving a thumbs up on Google. But simultaneously, people think this is the most important election of all time, and they are not tuned in. So it's like two-thirds of people think that it's an extremely high-stakes, you know, democratic moment, and one-third of people are paying regular attention to the news.
41:04So what's happening here? What's the fatigue? What's the distrust? And that's where, you know, we're building this new show in this environment. I'm so excited because Ben has had not one but two podcasts, I believe, before. right big big huge hits yeah that's why you don't know about them yeah i produced one myself back in city hall in 2006 there was a very faint beep buzzing in the background that i was never able to figure out so i just left it in but i interviewed elliot spitzer and rahm emmanuel people like that you know ben i did come up to you once in the buzzfeed uh newsroom and say i have an idea for a video podcast and uh you're like everybody wants a podcast these days and now we both have them.
41:49I know, and I gave people a hard time about podcasts, and I'm sure everyone I've ever told I wasn't that into podcasts is going to come out of the woodwork and mock me now, but I should say that now all I do is listen to podcasts. Third time's a charm, Ben, on mixed signals. Before I was setting up just this idea that Ben, you said that the media is a huge story, it's really a story of politics, technology, and culture, and then we have less media reporting now than we did previously and it should be a growing beat. I wonder why you think that is. It seems to me like a lot of media reporting sort of became like critiques of whether some publications were going to like call things lies and which ones would not.
42:33And that's sort of what it, it sort of became almost, I don't want to say partisan, but like it became small. Yeah. Well, I think we're in a moment of like peak media, like chatter and criticism, but with very, very little reporting. And we're easy to report on. Like reporters just leak so much, you wouldn't believe it. Like they're the easiest people to cover. They know the rules. They know how to play the game. And there is this really huge gap between this broad sense in all sorts of directions that the media is some kind of conspiracy with like wheels within wheels and the reasons you see things you see are because of something, you know, involving the Jews.
43:12and then the fact that you can then go and find out why things happen in media and it's not that hard and you can ask people. And it often is like some kind of bumbling conspiracy, usually a conspiracy to make money and a not very effective conspiracy to make money. But like things are happening for reasons which you can find out. And they're often not as on the nose as the reasons people imagine, but they're often really interesting. And I think like there was a story we did last week about Alito flying the flag upside down, which the Washington Post figured out and then didn't cover. And there was just enormous speculation, rightly, about what the hell were they doing.
43:50But it was not, and then, you know, we were the first ones just to call up the guy who made the decision and say, why'd you do that? And the answer was like, well, the Supreme Court reporter, who's sort of an institutionalist, wasn't that comfortable with the story. And, you know, like it was sort of very specific things, not a broad conspiracy. Although you can then say, well, that reflects the kind of, you know, change social moment. But in any case, it's a good story, the media. Recommend covering it. You know, I've always thought that the most interesting thing about media is that there's this unmonetizable portion of media, right?
44:23Like, I wasn't there. I wasn't there at the moment that Elon decided to buy Twitter now X for$44 billion. But I could tell you, with absolute certainty, Twitter at the time was worth more than$44 billion. It was just not monetizable for more than$44 billion, rise and fall of nations, you know, Arab Spring, propaganda, Russian disarray, I mean, elections, like the, the unmonetizable portion of media is where the power is. And like the gap between what is monetizable and what is valuable in culture is, is massive. And it, what makes media like, like out punch everything dollar for dollar, which is a question of like how you get the public interested in it.
45:04I mean, is there a business that can be found to sort of make money off this stuff if it's that influential? I guess that's been the question forever. Off of news? Off of, yeah, the non-monetizable news that Joe is talking about. I guess there's like this brand safety push, but I think, aren't you guys doing it in some way with events, like bringing people in? Yeah, I mean, I think, again, it keeps changing, I think. And right now, I think the people who are most successful, there's a lot of niche media that's very successful. I think there's better coverage. There's a company called Industry Dive, in fact, a really interesting company that provides really good coverage of a bunch of basically unglamorous industries that have never been covered that well in history.
45:45If you want to know what's happening in the waste management industry, which lots of people do, they do a great job. And I think there's analogies to that. Like Insomifor certainly is trying to cover a series of kind of verticals in a very deep way that appeal to the people who sit on top of them and who are your sources and your readers and also the people who are making decisions. And if you can reach them, there's an ad market that's not consumer, essentially, advertising. But you can still do news at a very high level that appeals to a lot of people. Because the sort of nature of news is that if you are breaking news that's interesting to the president of the United States, like a lot of other people want to know too.
46:31And so I think there's a generation of companies of which, you know, Axios, Politico is probably the first that have in different ways found that model. And I think there are elements of that in what we're doing. I think the question of like real general mass consumer news, like in a way what CNN is trying to do, very tricky right now. And it's so fragmented. And I think on the demand side, people don't necessarily want all their news from journalists. They want to get their news from, you know, Shannick McMahon or Heather Cox Richardson, right? And newsmakers would rather give their interviews often to these individuals as well, which, yeah, which is a real challenge, I think, for kind of mainstream journalism.
47:12I think it's the metric word, Ben, advertising. And the problem is like, you're right. B2B advertising is very different. Like, you know, if the president reads your newsletter or executives read it, like I know Alex has this with big technology, how do you monetize the fact that like there's 10 important people listening? Are you willing to pay a$500 ,000 CPM? Because the market has become liquid on this and Facebook, Google, Amazon, like everyone has convinced the world if you can't measure it, don't buy it. Right. So like all this immeasurable stuff is just out there in ether. Now, I tend to think that means there's massive inefficiencies in the market and we should be going trying to find those and buying them.
47:50But it's pretty hard to do at scale. And again, and again, because selling sponsorships is like running on a treadmill, whereas selling media is like being on a bike, like the bike keeps gliding and the treadmill, as soon as you stop running, you're going off the back of it. And I've spent a lot of time in the advertising side of the world. And I think it's, I think it might be the largest market inefficiency we have right now in the world. Okay, I want to ask one conspiracy question since, you know, there's talk about media conspiracies in your promo and we can sort of end on this. Do you think that on average, the media will have a rooting interest in favor of Joe Biden in the 2024 election?
48:29And how do you see that playing out in terms of influencing the election and the coverage that we're going to see over the next few months? I don't think there's any such thing as the media. That's why I said on average. There are outlets that have a rooting interest in both candidates, more openly expressed than ever. I mean, I do think there's this, I mean, and often when people say the media, they literally just mean the New York Times, which is how I take you to have meant that, basically. And like maybe sort of smaller subsidiary outlets like the Washington Post. The media on average. So add up all the news coverage, take the average.
49:02That's kind of like my perspective on this. Yeah, I think the average has stopped being relevant because consumption is so split. Like the average is a thing no one consumes, sort of sitting halfway between Fox News and MSNBC. I mean, in fact, that was where CNN was trying to live. I mean I do think that the kind of establishment media is like wrestling with the fact that it had the sort of years of arguments that whether if you really said that Donald Trump was a bad guy that that was the moral and ethical thing to do and kind of eventually came around to the point like yes we should say that Donald Trump is a bad guy in all our stories and then found that nobody cared at all when they did that and even the people who I think are sort of their reader base often found it kind of annoying And so they're now, and I think this is sort of a challenging timing for Joe Biden, the places like CNN and the New York Times are sort of in the middle of these internal culture wars that are about pulling away from a sort of more partisan stance and away from sort of more left-wing cultural politics.
50:02And that's happening just like simultaneous with the presidential campaign, just slightly by coincidence. And I do think that if you're in the White House, it's like, huh, like maybe you could have like postponed your purge of the partisans for like four more months. But right now the priorities inside the times are very much about reestablishing their independence from among other things, the Democratic Party. The show is Mixed Signals. You can find it on your podcast app of choice. This was a super fun conversation. Thanks for coming on, Joe, Ben and Naima. Hopefully we can do it again soon.
50:32I'm looking forward to listening to your show. Thank you, Alex. Joe, We look forward to talking more about our AI project. Let's do it. You think Joe's just Mr. Deep Pockets now. Alex, do you have any tips on working with Ben? I mean, Ben's the man, and he's got a great intuition. So wherever his intuition's taking you, just run with it. Alex had the worst hiring process in the history. It was so terrible. Have we talked about this enough, or should I bring it up again? No, we can talk about it again. I met Alex, and he was like, I really wanted to hire him. He was on my list of really good reporters.
51:03but also I had like scrambled his name with somebody else's. And he was like in the hiring process. And I called him out of the blue and was like, hey, do you speak Russian? I needed a Russian speaker at that moment. And he was like, what? Right around the Boston bombing town. No. And I just hung up on him. And then I didn't get the job. And they didn't call back for like years. It was terrible. Chaotic, Ben. I mean, yeah, I think about a year and a half later, BuzzFeed ended up hiring me. Yes. And you spoke Russian. And Ben brought everybody around. I was like, oh, maybe I should have. You learned Russian in the meantime.
51:36I know, spasiba. But Brent brought everybody around in the newsroom. It wasn't even my first day. It was like, come and see everybody. And he's like, hey, listen to this story about Alex speaking Russian. But I will say, it took three times to try to get in. But man, five amazing years within BuzzFeed and really taught me how to be a reporter. So thank you for hiring me. Maybe Vivek is going to bring both of you back, guys. Maybe you're going to have a reunion. I am so ready to publish those drafts. Yeah. Yes. He needs some right by the influencers. New movies is the thing he ordered. Exactly.
52:11All right. Good seeing you, everybody. Thanks so much. We'll see you next time on Big Technology Podcast.
From the publisher
Ben Smith and Nayeema Raza host Mixed Signals, a new media podcast from Semafor. And Joe Marchese is the general & build partner at Human Ventures. The three join Big Technology Podcast for a VC + Journalist conversation on whether the news industry can survive its dealings with generative AI tech providers, like OpenAI, after several publications inked multi-million dollar content deals. Tune in for a conversation that explores whether the news industry is setting itself up for a repeat of past mistakes made with Facebook and Google. We also cover whether the news industry can harness AI, what the value of news is to AI companies, and whether AI can be taught to emulate media brands and pay a licensing fee. Plus much more. Tune in for a lively conversation about a core issues facing the AI field today.
---
Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice.
For weekly updates on the show, sign up for the pod newsletter on LinkedIn: https://www.linkedin.com/newsletters/6901970121829801984/
Want a discount for Big Technology on Substack? Here’s 40% off for the first year: https://tinyurl.com/bigtechnology
Questions? Feedback? Write to: bigtechnologypodcast@gmail.com


