The growth and proliferation of artificial intelligence is causing many industries to reimagine what they will look like in the near future. The authors of a new book believe that is certainly true for journalism. Rodrigo Zamith, Tomás Dodds and Seth Lewis are the authors of a new book called "Journalism in the Age of AI: From Acceleration to Reimagination." Zamith joined me over Zoom earlier this week to discuss the book and his research. He says he has previously written on the impact of AI on the news production element of journalism specifically.
Zamith: But very quickly, as I was working on that book, AI started to really evolve, right? With the public release of ChatGPT. All of a sudden, I became far more public facing in ways that became tangible to more people. They were using it directly in their lives. And it became very apparent that this phenomenon, writing about it solely about journalism, was missing the point. And so we decided to, my co-authors and I, Seth Lewis and Tomás Dodds, to work on a book that focused on more than just journalism.
Journalism is a starting point, but then it's an avenue to also look at how AI is impacting digital infrastructures, right? So the way that people get information, about how it affects how people are processing information, right? Instead of seeking links to news articles, people are just seeking answers through AI overviews, right? And then more broadly, what it means for educating people. So not only how are people processing information, but what skills and critical knowledge do we need to be imparting in our students going forward? And even how do we need to be changing the assignments that we offer in the classroom?
So I think when people approach AI through the lens of, is this going to just be replacing journalists? I think that is too narrow of a question.
Moore: When we think about technology and the kind of tools we have now, when we think about them in their early stages, whether that was the computer, whether it was the internet, whether it was digital editing software. I mean, I'm using a digital editing software right now to record this conversation. I'm not doing it on literal physical tape. I'm doing this digitally. Right? When we think about the early days of these sorts of technologies and look at them in parallel with where AI is or AI is going to be, what are some lessons that we can learn about not necessarily viewing it as a, you know, a get out of jail free card. I don't have to do anything critical here anymore, and view it more in the way that we use computers and the internet and digital editing softwares as a critical tool to doing our jobs.
Zamith: I guess with AI, what is different again, is the fact that since it can do so much for you, it is easy for you to turn over more of that work to it. But if everyone else is turning over more and more of their work to it, then what you're getting is more homogenized work, right? More work that is very similar across different people. And if I'm a media producer, for example, then I start to wonder, well, what is distinct about my work? What is different, right? What is valuable? Why would people want to come and hear from me or consume my work, as opposed to somebody else that has access to these same tools and is relying too much on it?
So I think for me, with these technologies, it's both. It's two things really. One, it's again, how do I maintain my own originality? Meaning how do I not let the allure of having to do more of the work for me not get to a point where I'm offloading so much that my work is no longer really original? I think that's one piece.
And the other piece that connects to earlier technologies that you mentioned is that they all promised to liberate us from work. But most of the journalists that I talk to don't feel like they're working any less today than they were two decades ago. If anything, it's quite the opposite, where they feel like the expectations are that they can do more, and therefore they should be doing more, producing more stories and interviewing more people and so on. And I think that's something that we need to be cautious about with AI as well.
That's a big reason for the subtitle of the book, from acceleration to reimagination, which is that this is a technology that can really accelerate the work that we are already doing. We use the metaphor of a hamster wheel in the book, right? Where a lot of journalists, they have been asked to run faster, but ultimately feel like they're still running in place. And we hope that AI can be a moment to reevaluate, you know, what are we doing? Right? Can we be stepping off of this hamster wheel? So we're not, again, buying into technology that's going to make us do more, but not really do anything different. Arguably maybe do even worse work at times. Or can we use this as a real pivotal moment to just say, what can we do to differentiate ourselves when more and more power is being taken away from journalists and media producers and being transferred over to intermediaries.
Moore: One of the things that's changed very significantly over the last quarter century with journalism has been the economics of it. How do we pay for it? We've moved away from classifieds and really, we've moved away from a world where Google is a productive way for us to see engagement in that way to, we've moved into a world of Google Zero where there's no more going to Google, typing in something and expecting someone to go to your website. That just doesn't happen anymore. So when we think about all of these factors in place, what impact do you believe AI will have on that changing economic landscape for better and hopefully for better? Not necessarily just for worse.
Zamith: Yeah. I mean, I'm going to be completely honest with you. This is the piece that I'm most pessimistic about, right? As you mentioned, the economics of journalism have been difficult for quite some time now. And when people think about AI, when I have conversations about people, they are worried more about how it's going to lead to direct job losses for journalists, because AI will replace those journalists. And I don't think that is the biggest question. To me, that is the biggest worry. To me, it's more about how AI is going to be changing the fundamental economics supporting journalism. So not whether, again, the synthetic agent can do the same work as a human being at the same level of quality. To me, it's more about whether it is simply going to undermine the economics of journalism so much that we are going to have less and less journalism, especially at the local level, right, where we have seen journalism suffer the most.
And so, yeah, you mentioned Google Zero as this phenomenon, where in the past people might have gone to search engine and typed in a question, and hopefully Google would surface the most relevant articles to them, they would click through to that article, which meant that the news organization would at least get some ad impressions out of it. Maybe if the content is good enough, it might get some people thinking about accessing more of their content, so more ad impressions, or maybe eventually becoming a subscriber. That was always the goal.
Now with AI and specifically with chatbots and AI overviews, people are less likely to click through to the original content. And that just means that more and more of the value is being extracted from news organizations by tech companies. And ultimately, it's creating a bigger gap between the news organizations and their audiences, because you're now inserting this intermediary that is oftentimes answering the questions that people have. And that's, I think, another scary thing for news organizations, where I think we are finding that people more and more are willing to settle for answers that are convenient, that are quickly accessible, that might not be fully complete or fully contextualized, and in some cases might have some factual errors, but that the convenience and the ability to directly at least make a person feel like their question has been addressed. That is enough for most people. And again, that impacts the economics.
And so when I think about the future of AI and journalism and how do we incentivize using AI in a way that is beneficial to journalism? I don't think that necessarily starts with news organizations and what news organizations and journalists are doing. I think it also very much starts with citizens. And what is it that we are willing to settle for, and what is it that we're willing to reward and demand, again, not just from news organizations, but maybe from our regulators as well? Right. I think right now, the incentive structure for AI development is kind of perverse and is not rewarding what is good for democracy. And so it's on us to demand something different and to reward something different. If we're unhappy with that trajectory, which I can say I am not.
Moore: The book does a great job of looking at the 30,000-foot view of how journalism at large will be impacted by the growth of AI. Here on Ozarks at Large, we're a team of less than a half dozen people who mostly work on the show. We do a lot of other tasks around the radio station, too. When you think about newsrooms, whether it's the one that I'm working in or one similar to ours, what advice do you have to newsrooms of this size or local media outlets who are dealing with making sure that they continue to stay relevant? They continue to offer good local news that only they can offer? And how do we think about the use of AI in our work?
Zamith: Yeah. So I teach at the University of Massachusetts Amherst. We're also in a quite small town, and we have noticed that there's just less local news coverage right now. So if I'm a local media organization, I'm thinking about AI in terms of, are there already recorded public deliberations, for example. So city council meetings or town council meetings for us that we can use AI to process transcripts from those public deliberations and automatically create a list of topics that were discussed at that meeting that might not have been on the agenda, for example, and to start suggesting story ideas for us, especially story ideas that might connect to previous work that we have produced as an organization. This is all possible using existing technology. Some of these tools would even be free. Others would involve a freemium tier or perhaps a paid tier, but they're still quite affordable even for local newsrooms.
And for me, that becomes a way of just saying, in the past we might not have had enough people to send to these meetings, but at least now we can have a watchful eye that is suggesting ideas that will not lead to AI writing a story about the city council meeting, but instead is just helping us identify what are some things that might have been talked about there that could lead to a relevant story for our audiences. And so it's a mixture of asking for story tips to figuring out what are some of the most interesting quotes that came out of there? Right. Are things that were said in that meeting incongruent with things that might have been said in the previous meeting, which again, maybe you're sending a journalist to cover the meeting on a Thursday, but they couldn't go in the previous week, so they might have missed that incongruence. Right.
So to me, the advice I would offer is to, where possible, set a little bit of time aside for experimentation, right? Don't approach AI through the lens of this is going to solve all of our problems, or don't even, I would argue, feel the need that you have to be using AI. But I think just being open minded to the fact that this could be something that is useful to you, and be open minded about learning. I think that's where most people figure out how can it be useful to us in our specific context for the audiences that we're trying to reach, and maybe most importantly, for the work that we're trying to do and the values that we have as an organization?
Moore: Rodrigo, it's been a great conversation. Anything I missed? Anything make sure we touch on here?
Zamith: I think maybe one of the final things that I hope people take away from all of this is that the trajectory for AI is not set in stone. I think AI is going to be a part of our future, but if we are unhappy with the way that AI is being developed right now and implemented, then I think we still very much have agency to speak up for what we value, to reward the people and organizations that are doing things that align with our values, and to ultimately demand that the organizations that are using AI in negative ways be held to account. The future is not set in stone, and we still have a voice, and we should be part of imagining that future and making sure that we are able to realize it.
Rodrigo Zamith is an associate professor in the journalism department at the University of Massachusetts Amherst and a co-author of the book, "Journalism in the Age of AI: From Acceleration to Reimagination." We spoke over Zoom earlier this week.
Ozarks at Large transcripts are created on a rush deadline and edited for length and clarity. Copy editors utilize AI tools to review work. KUAF does not publish content created by AI. Please reach out to kuafinfo@uark.edu to report an issue. The audio version is the authoritative record of KUAF programming.
