EP 455: AI Is Painful for Journalists, But Good for Journalism

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

I'm not a robot. I don't know why sometimes a lot of people accuse me of being AI and say, oh, you're a robot. I'm not, right? I'm a real boy. And when I was a real boy, like quite literally, I was a teenager, I was a journalist. I spent about the first seven or so years of my career in journalism. And now as someone that literally talks about and covers AI every single day, I'm always torn, right, because I feel for the journalism industry and a lot of the ups and downs that I think artificial intelligence and large language models puts the journalism industry and journalists through. But there's also probably some huge bright sides. So I'm excited to tap into one of my old professions and to bring on a real professional to help guide us all through it. And we're going to talk today about how AI is painful for journalists, but maybe good for journalism. All right, I hope you're excited for this one. I am. And if you're new here, welcome. This is Everyday AI. So this is your daily live stream, podcast and free daily newsletter helping everyday people make sense of the mess of AI, right? It moves so quickly. What does it mean? How does it impact us all? That's what we do here. So welcome. This is your home and your second home is, well, your everyday AI dot com. While you're there, you can go sign up for our free daily newsletter. We're going to be recapping today's conversation and a whole lot more. We always keep you up to date with the latest AI news, tips, tutorials, all that good stuff. But we're going to be recapping today's show. And while you're on our website, there's more than 400 episodes. You can go listen to them all for free, watch them all, read about them all on our website, sorted by category. No matter what you care about, it is there. It's like a free AI university. All right. And if you are tuning in, we normally go over the AI news. Technically this is pre recorded one, debuting it live. So make sure to go check out the newsletter for that info. All right, enough chit chat from me. Let's bring on the real experts, shall we? So, hey, live stream audience, I'll, I'll still be in the comments, so you know, leave, leave us questions. I'm sure me and our guest today will go through and answer them. So happy to have on the show and help me welcome. There we have him, Pete Pashal, the founder of the Media Co Pilot. So thank you so much, Pete, for joining the show.

Hey, Jordan, thank you. Pleasure to be here.

All right, I'm excited for this one. Get to bring in my old world, but tell everyone about your current world. What do you do at the Media Co Pilot?

Cool. Yeah, well, the Media Co Pilot, I founded it about a year and a bit ago and it started out as a newsletter. I challenged myself to learn everything I could about how AI is changing media and journalism and pass on that knowledge to my readers. And soon after that, some folks started reaching out to me and said, hey, can you come by to my team and like, explain this AI stuff? And I was like, I could do that. And so I did that once or twice and that turned out to be really a really good experience. And that's why I sort of launched formally AI training in 2024. And I've been teaching journalists, PR teams, creative teams, some people far afield of that even about how to use AI in their work and do a little consulting on top of that. So, yeah, it's been good times.

Love it. So, you know, walk, walk me through some of the. Some of your early trainings, maybe. So you walk into a room full of journalists who, you know, if you're not from the journalism industry, the whole whole job loss thing has been around for many decades, right? Like when I started in 2000, what year was it? 2002 or something like that, everyone was scared of the Internet and, you know, it's going to take away jobs. And it did, I think the advent of the Internet and the availability of information. But, Pete, walk us through. When you walk into these newsrooms needing to teach AI to a bunch of journalists, what, what are their questions? Are they excited? Are they scared to death? What's it like?

So I'll say this. About a year ago, it was much easier to blow people's minds. You know, you could walk into a class and show sort of some basic prompting with chat, GPT and a little bit of advanced stuff with how it could manipulate data and things like that. And everyone would walk out with, you know, I'd see the light bulbs going off and they would just be like, oh, I really want to apply this and do that. That over the past year has. Has definitely waned. People are like, okay, I get it. You know, this thing can create content out of whole cloth. I could create videos, images, this text. I've Actually prompted a bit, you know, show me something better. So I've had to gradually raise my game throughout the, throughout 2024 and you know, I've had to essentially really upgrade things with advanced techniques on, you know, using the AI for PR and media monitoring. In a newsroom, it tends to be partly about the journalists, but also about the operations. You know, how do you get this stuff in applied in a systemic way, particularly if you don't have a ton of money to invest in like a new cms. Right, because no one wants to like migrate and that sort of thing. And that sometimes involves showing some more advanced stuff. There's a lot of like no code platforms for example. So if you want to go beyond just copying and pasting from a chatbot, it's actually relatively easy to design your own automations to connect with the chat experience or the AI experience so that you can basically have a trigger like an RSS feed, for example, and then the content goes to the LLM and comes back to you in some form and you can do that for multiple workflows. That's where things have gone from. Okay, it's cool that we can do content. Awesome. And to how do I do that to actually save time and redirect the way I work? And that's been probably the biggest shift.

So, you know, could you maybe walk us through more from. So, you know, you kind of just explained some of the thought process and the working relationship, you know, between you and people you're trying to teach. How would you say the industry in general, and I know this is a hard right to speak for an entire industry, but how would you say the journalism industry in their perception of AI and has changed even in the last year?

Yeah, really good question. So I would definitely say that there's a lot of inherent skepticism about AI from journalists, probably more so than a lot of other professions. While everyone's excited to unlock sort of the efficiencies and you know, everything from the low hanging fruit to advanced stuff, there's sort of an existential quality to journalists and writers broadly. Right. Because it's. Well, this AI can now essentially do at least a simulation of my job, like the writing and the editing and there's an uncanny valley aspect to it certainly, but it's like, oh, I don't know if I like this. And I think that's while simultaneously having your mind blown, it's kind of seeing sort of your job replicated in this robot way is very unsettling. So I think that still exists to a large Extent, I think it's waned a bit as the reality of AI set in. I think, you know, there was a lot of talk and there still is to some extent, some skepticism about whether this is a bubble, right? Well, it might be in some business senses, but like the Internet, which, yes, was a bubble, but also was a super long term important trend. I think even, even if the bubble bursts on AI, AI is not going anywhere. And there's a reality that it's, it's just going to change the world, journalism included. And perhaps especially so. Unfortunately, AI is coming at a time when media is challenged for other reasons. There's a sort of natural cycle to add ad markets, and we were definitely on the downside of one of those for a while, but also just in the larger timescale of the media itself. Like the 2010s were filled with a lot of what I call scale media. Right. So search and social media rose up as these amazing refers like Facebook traffic was off the hook for so many publishers for so many years. And searches, you know, grew into something that everyone sort of depended on and still is to a large extent. But that wasn't really that great for journalism. You know, there was a lot of like, not great stuff written. And, you know, there. I used to work at Mashable for a long time and one of the more popular go to formats for quick hits. And by the way, the whole idea of quick hits is a symptom of this era. But one of them was like a go to format of let's do, let's see what people on Twitter are saying about this thing in pop culture or in business or in tech or whatever. And you just get a bunch of tweets and compile them and call it journalism. And it's like, I'm sorry, I hate to break it to everyone, including myself. I've been guilty of writing those things. That wasn't journalism. That was not great. A robot can do that just as well now, and a robot might as well do that. So, you know, like, so, so there's, there's a certain weeding out of content and maybe people who, who were doing that content. But at the same time, there's a broader, like, let's get leaner kind of ethos. And this is, again, this is not restricted to media and journalism, but because it's coming at a time when I think people were expecting a pendulum swing back to like, oh, the ad market's better now, and maybe people start hiring again instead. Most companies are like, wait a second, let's maybe not start hiring again. And let's just make sure we're using AI to the utmost before we get a bunch of heads that we might have to say goodbye to in a few years when things swing back again. So that's kind of like the, the, in a nutshell, what I see as the, the pain in the journalism industry. But what remains? So, like, I, you know, I still think, and I, I'm optimistic about journalism itself in the AI era because, because we're starting to throw out those incentives of search and social, which, like, as I just sort of explained, we're probably not that good. These dopamine spikes you get from traffic spikes, not good. Even though, again, they still exist to some extent. But AI now adds this ingredient of summarization. And so there's these disintermediary sort of interfaces. So chat GPT is one of them, perplexity is one of them. You know, there's, there's, honestly, like, even news organizations themselves are starting to put bullets at the top of their stories that are generated by, by AI. And, you know, Amazon's going to start having Alexa summarize the news for you, et cetera. Like, you know, there's this layer of summary. This is how a lot of people are going to start getting information, which means that you're going to have to start optimizing for that experience. So what is that experience looking for? Kind of, you know, is what the question is. And honestly, like, I'm going to, you know, lay my cards on the table. I don't have pocket aces. Like, I don't know either. All I see is patterns from what I see these things doing and what people in the industry tell me and their impressions of it. This is being studied, by the way. Obviously, like SEO for chatbots will be a thing even with its different incentives. But so far I can see AI, again, assuming it's designed well, which is a big caveat. You know, it's going to prioritize thoughtfulness, it's going to prioritize stuff being comprehensive, stuff being unique and stuff being what I sort of, this is my term really, but it's like definitive, like something that if you left it out of the summary, you would be wrong to do that. Right. So that's kind of the content people are going to want to produce. And I would say that's what we used to call good journalism.

Yeah. You know, and so one thing my mind always goes to is the monetization of this. All right? Like, it's, it's always Been in any, you know, especially I came from newspapers. I'm sure it's different, you know, depending on if. If you're at an online publication or, you know, media, broadcast media, whatever.

Right.

But everyone's always worried about, right, the clicks. You know, especially in the early days of online ads, it was the clicks. You know, you talked about your experience in Mashable. It's the clicks. It's the time on site because that brings in ad revenue, right. Which pays said journalists. What happens next? Because what I'm seeing and what I've been saying for a while is it seems like the Internet's getting harder to use because publishers and big content companies are losing clicks. So they got to double down on their, you know, on their display ads. And it gets hard to read that news story. Right. Like, I've taken screenshots, Pete, I kid you not, on big, big websites. And I'm like, the entire screen is ads and I, like, I can't even see the story anymore. Right. So what. What happens? Because these media companies need to keep feeding the beast, right?

The.

The big tech companies, big AI companies that are essentially just scraping copyrighted information, putting it in a bowl of spaghetti, spinning it around and saying, oh, it's. It's an Italian dish. Now it's something else. Right? How's this thing gonna keep going?

Yeah. Wow. You really soaking a lot of, like, directions I want to go in, but, like focusing just on the monetization way to like, kick at the bad knee of the media. That has been.

Yeah, I have to call a spade a spade here. I. I have.

Right?

I have. I'm cheering for both teams, though, Pete. I am.

No, that's totally right. And this is really important to talk about. Yeah, they're like, I. A lot of the phenomenon you're describing is encapsulated in, you know, this is this mega viral post from. I think it's about a year ago now. The Cory Doctorow is in shidification of the Internet. And I hope. I. I'm not sure if I can say that you have to bleep that out maybe for the long term, but it's totally right because all of these incentives around content, which I'm going to be clear and point the finger, like Google created essentially, not solely Google, but it is the biggest player in this in terms of doing things like forcing recipes to have a whole bunch of text that is kind of meaningless and then. Then that jump to recipe thing, you know, for a long time. I remember we used to play the game this was more than 10 years ago now of galleries. You would just click through galleries to create more ad impressions. People eventually got wise to that and learned it was a trick. But the overall sort of CPM Internet is still exists to a large extent. So the big question is like, what does AI do to this? Right? And so far, again, let's be clear, it's early days for AI and the monetization systems. Google has only just recently in the last couple of months started putting ads in their AI overviews, their summaries and that. And that's probably just the first move, right? Like, because that's just like an ad strip that they put everywhere and that the idea of having placement within the summary in a different way, in the same way that they do for search results, right, all the sponsored links, something like that is probably coming. That said, they got to tread a lot more carefully because you have these cleaner interfaces over at Chat, GBT and Perplexity. That said, you know, ads are coming in. They already have, like Perplexity's finally done its ad platform. Sam Alton speaking openly about ads being an ultimate model. But they're well funded right now, so they can sort of keep things clean and sort of bring users in. So hopefully that can be preserved to some extent. But if it does stay preserved, then it's like, well, how do you monetize? So far the, the, the answer seems to be, at least on the publishing side, licensing. So you're not getting clicks anymore because now you're getting summarized, you're getting substituted by this AI summary on some platform somewhere. That platform is going to have to pay you some amount for the privilege of doing that. Now, right now that is a big legal gray area because for the longest time these big data sets that Google was using to crawl the web and others were using for research, you know, it was basically fair use. It was like, we're indexing your site, therefore you're going to get traffic. And this is a grand bargain. Well, now that bargain has been severed very much on one side and it's kind of like, well, wait a minute now, we don't want you using that content. In fact, it's funny, this is reported in Wired Common Crawl. One of these big data sets had not had anyone really ask for their data to be removed from their data set until the last couple of years, until everyone suddenly woke up and was like, oh, AI is doing this and they're not giving me anything in return. You know, so there's obviously been all these lawsuits. You know, the New York Times is suing OpenAI. News Corp. Suing Perplexity. And, you know, the, the I, the idea at the, when the New York Times sued OpenAI was mostly about training data. And these search engines were, you know, Perplexity was still pretty nascent. And ChatGPT, while it could access the web, it was very clumsy and it wasn't really a search engine. Well, that's a different world now. And now I think AI for a lot of people today, as opposed to a year ago, is now more about doing something with data, or, you know, in search results are data, as opposed to just like I want something from your knowledge base, from your training data, and give it to me, hallucinations and all. So because of that evolution, that lawsuit looks a little obsolete now. You know, that, that old one, because it was more about the training data and outright plagiarism of OpenAI. But the point is, the consensus of the industry, with all these media companies signing deals with OpenAI and others, is pointing towards the use of AI to summarize media content is becoming much more akin to syndication than it is about fair use or aggregation. You know, and maybe we'll get a ruling on this, maybe we won't. We'll see. I think some of these court cases might be settled. I used to think the New York Times case wouldn't be settled. I thought they would be on principle and they'd go right to the end. But because of that obsolescence, I just mentioned that it really focused. Like if you look at that filing they did, it focuses so much on people bypassing the paywall of the New York Times with this duplicate of content. That's just not a use case anymore. It never really was, but it was like they've kind of weeded out the plagiarism thing mostly from these engines. I just never hear about it anymore. So I feel like unless they radically change their argument, and there are other arguments in the lawsuit, it's just they focus a lot on like, oh, this article was completely copied. But unless they do it, I think they'll probably some kind of settlement, which is too bad, because I do think there should be some kind of ruling. And the simplest one would be to say, like, okay, training your AI on publicly available data is okay, like, that's fair use. But using AI on someone's data in a, in a sort of essentially a rag way, or like summarizing news way, that requires a business deal, and that's not okay. So I think, I think that's the simplest way to break it down if I were, you know, if I were a judge trying to speak to the future and not overcomplicate things.

So, you know, it's interesting. I'm glad you brought up the New York Times versus OpenAI case. I'm going to try not to go on a random rant because I did like an hour long podcast on that when it first came out. What was that? December 2023 I think it was. But even before that, at the time I said this was early 2023, I said big, big media corporations, big, big news organizations. They're either going to have to sign into partnerships within, you know, OpenAI, Google, whatever. They're either going to sue them. So you either partner with them, sue them, or you maybe just go out of business. Is that too harsh or could that be a reality for journalism of the future? You choose one of the two paths. If not, good luck to you. Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everydayai.com partner to get in contact with our team or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on Genai.

Yeah, I hope not. I do think there's some, some hope in that. AI isn't the be all and end all on all information. I do think even in using AI search, as useful as it is, and I talk to ChatGPT all the time now on my iPhone. I still want that page of blue links sometimes, not all the time. But I'm also a power user of the Internet and I sort of have a rhythm of, you know, the processing that an AI often does on a search. To give you the answer, I can actually do in my head a little quicker by looking at that page, you know what I mean? And I kind of want those links because sometimes the correct link doesn't always come up first as quickly when you do a search. But that's me. But I think for the media and audience broadly, it's a matter of like, if you want to be in an AI search. Right. Like, I sort of was articulating that earlier. I would say the question to answer is like, why would you want to in the first place? And if you have either a plan to get licensing around that idea, basically, is there an ROI for you on being in that search, then great, do it. And if not, maybe not. But then you really have to think about what's your growth strategy, what's your audience strategy, which is important anyway, because AI is the only thing. But it's, it's just a matter of like, are those vectors, those other vectors going to perform for you? You know, that's a, that's a tough call on, on a lot of media companies, but I think it comes down to how differentiated is your content, how in demand is it, how can you, how much can you force the AI companies to come to you? Now, if you're, you know, currently they're not interested in smaller publishers. There are a whole bunch of startups that are trying to solve this problem by creating marketplaces. So like a smaller publisher will go to the marketplace, put their content up there for a price, and then on the other side a AI engine like Perplexity could go and like, I want, I don't know, crypto content. And so there's, oh, we have a dozen people, publishers putting it on here and you can maybe pick and choose depending on the price you want to set. That's the, I, that's the dream. That's a scalable thing that everyone can sort of, sort of probably get behind if it works. But at the same time, you know, that's, it's a, it's very early days and there's so many of these things now. There's like about a half dozen of them. At least one of them, I think tollbit is sort of starting to come out ahead because they have some real, some deals with real publishers and they've gotten some decent funding, but we'll see. There's, there's a bunch of them out there that are trying to do this thing.

Yeah, I like, I like that future. Right. Like the, the, the Uber for, you know, personalized media publishing or personalized media consumption. I love it. So, so, so far we've gone in depth on A lot of things and there's so many things that I haven't covered that I want to talk to you about. So without turning this into a accidental, you know, Lex Friedman three hour podcast. I'm gonna see if we can go kind of quick bullet point here. So let's, let's see. So Pete, I'm gonna ask you a couple questions and let's just go kind of rapid fire speed round.

Love it.

Let's see, let's see. Let's end it with the speed round. So when we talk about AI's impact on journalists, right, we said there's pain. What are those biggest pain points? Give me that, that bullet point, top of the page. What are the biggest pain points for journalists when it comes to AI, I.

Would say it's basically you, you need to one, upskill yourself, you know, because there's, there's the person in the desk next to you who's going to be the AI qualified person who's actually using it for research to get reports emailed to them, first thing to scrape sites and to do a lot of summarization for their story building that you might not have. So, you know, you kind of need to embrace it just to kind of like keep pace. And like I said, if you are in one of those positions, probably a junior role where you might be doing sort of content where you're not really adding much, you're more summarizing something that happened and getting a rote comment from an official that sounds like a job destined, at least 80 to 90% of it to be done by a robot. And then you're going to be tasked with just kind of like putting stuff around the robots work. This is a real hard problem because there's the pipeline issue with AI of like, what do the interns do, what do the junior people do when AI can probably do their job a little better? And it's kind of like, well then how do you get senior people if you don't, if you let the robots do their work, right? So this in the long term is a real existential problem, not just for media, but I think essentially a lot of these jobs are going to have to be reimagined either a bit or completely to make sure the human work, mostly in journalism, it's sourcing and putting things together in an insightful way that that is emphasized throughout, from top to bottom, all from interns all the way to the editor in chief.

All right, and then on the flip side, give me the bullet point summarization. Why can AI or how can AI be good for journalism?

The bullet point summarization of that is that AI will essentially change the incentive. So it's again, the other side of the coin I was just talking about, which is that, okay, so we're prizing unique content insights, the original, sourcing the new information. Well, if AI, if an AI mediated ecosystem rewards that, you've got to do that as a journalist. And like I say, that's what good journalism is. Go out there and call people, go out there and make sure you're getting stuff that hasn't been reported before, stuff that someone doesn't want published. Someone told me that was a great definition for journalism. I, I completely agree. And, you know, did you basically bring that hard work? This is the hard work of journalism. Right. Again, this is why I'm very critical of the search social era. It was so easy just to do a hot take on something and post it on some network and watch all the Facebook traffic roll in, especially if you gave it a provocative headline that was sure to bait someone. Right. Now, again, assuming the AI is designed properly and doesn't reward that bad stuff and rewards the, the good information, that's what you're going to have to produce. And it's, you know, it's harder, but it's a good hard. You know, choose your hard is sometimes is a sort of a viral thing and social media, it's hard to build a career on just social junk because it's a house of cards. But it's also hard to do a real journalism. I would choose that hard because you're going to have a much more solid foundation.

Great takeaways. Love that we were able to look at both sides of the coin today, Pete, on a topic that's very near and dear to my heart. So thank you so much for taking time out of your day to join us on the everyday AI show. We really appreciate it.

Oh, it's my pleasure, Jordan. Thanks so much for the stimulating conversation.

All right. And hey, that was a lot, y'.

All.

Maybe, maybe you missed some of those golden nuggets that P was dropping on our virtual heads. No worry, we're going to be me, a human recapping it all. I'm going to put these fat fingers on the keyboard, type this up in our newsletter as well as everything else that you need to not just keep up with everything happening in the world of AI, but how you can get ahead and be that smartest person in your company when it comes to AI. So thank you for tuning in. Make sure you Go check that newsletter out at your everyday AI.com if you're listening on the podcast. Thank you. As always, make sure to please subscribe to the show. Tell your friends. Don't just keep this as your secret cheat code. That's rude. Don't be a jerk. Share it with others. Thank you for tuning in. Hope to see you back back tomorrow and every day for more Everyday AI. Thanks y'.

All.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit your everyday AI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.