Our podcast: What should we expect from journalism in 2024?

"It's a really important year to help define how journalism and AI are going to work effectively together," says Nic Newman in our podcast
Trends and predictions social card
19th January 2024

The topic

2024 promises to be a significant year in journalism. The rapid rollout of generative AI presents new opportunities to enhance journalistic processes but also potential risks. The decline of the platform referral model is forcing newsrooms to look at building more direct links with audiences using a wider range of multimedia formats. And a host of world-changing news events keep newsrooms asking themselves the best way to engage core audiences without alienating others. We discuss all this and more with the author of a yearly report tracking the fundamental trends shaping journalism and how news leaders plan to negotiate them in the year ahead.

The speakers

Our guest is Nic Newman, Senior Research Associate at the Reuters Institute and lead author of the report Journalism, Media and Technology Trends and Predictions 2023.

Our host is Federica Cherubini, Director of Leadership Development at the Reuters Institute.

The podcast

Spotify | Apple Podcasts

The transcript

An uncertain world | The end of mass referrals | AI and journalism | Perceived risks of AI | Plans to tackle news avoidance | The use of different news formats | Revenue raising strategies | Next generation AI

An uncertain world 

Federica: Can you remind us what this yearly report looks at and how you went about it?

Nic: Yeah, I mean, this report is forward looking, it’s about trends. And it's really trying to identify the key audience, business and technology trends and sort of bring together what we know, how all those things are likely to impact on journalism. So that is everything from rising news avoidance, to advances in AI, and how they're affecting content and the way it's produced, the way it's distributed. And this is really looking at analysing our own data, other people's research and data. And also talking to industry leaders. So you mentioned a survey we do with more than 300, editors, CEOs, heads of innovation. But we also talked to some of those in depth as well to find out what's on their mind.

Federica: And what's the mood for the newsroom leaders out there?

Nic: I mean, it's mixed, I would say, you know, the key word is uncertain, less than half, So 47%, said that they are confident about the year ahead, the prospects for journalism in the year ahead. And the rest have various concerns ranging from declining advertising revenue, the fact that cookies are on their way out, which may make that worse, the business side, I mean, declining news consumption, news avoidance, I've mentioned that already. And of course, in some parts of the world, it's much more about just survival, how you can be independent, when there are increasing attacks on journalists, those may be physical, legal, or verbal. So a huge range of different problems. It's hard to sort have a single problem, but I think, compared with a couple of years ago, people are much less confident, because it's a very uncertain world, on the business side and on the, you know, just how politics is as well.

The end of mass referrals 

Federica: If we look in depth at of some of these issues. So media leaders you surveyed say they are concerned about the level of audience referrals from social media platforms. What are they worried about?

Nic: Well, pretty much all the publishers we talked to said that they'd seen a really big decline in referrals and traffic from what we might call traditional social networks. So you know, Facebook, primarily, but also Twitter, and others, in the last year, and we saw some data from Chartbeat, which is a data analytics company, which showed 48% decline in referrals from Facebook alone, and 27% from Twitter, or X, as we must now call it. And that's based on around 2,000 publishers in their network. So it's kind of aggregate data. So it'll be different for different publishers. And the reason for that is that Facebook have been quite aggressively pulling back from news, partly because they're trying to focus more on creators and content that is perhaps a bit more uplifting, connecting people to other people, rather than news, which can be depressing.

And then, of course, you've had Twitter with all those changes in policy and upheavals in the Elon Musk era. And that has also had an effect on, how valuable the user base finds it for news, but also what publishers can do there. So about two thirds of people in our survey of respondents said they were really worried about this, not just about what happened last year, but what might happen this year in terms of further declines. And this move to essentially, to engage people in social networks now you have to create bespoke content and often video content that costs a lot more money in networks like TikTok and YouTube. And that matters because many publishers have depended on this referral model to provide a big number for their ad business, or even if you're a subscription publisher, top-of-the funnel interest for people who you might then want to approach with a subscription offer. So, this is really, pretty existential, if it continues in that way it does, then publishers need to find alternatives

Federica: And of course, as you said, those more traditional social media platforms were also based on a link referral strategy and some of the new platforms, maybe less. So what are the publishers planning to do about it in the landscape that you just described?

Nic: I think the overwhelming focus is on direct traffic, can we build direct relationships? So if we can't rely on these platforms to provide traffic, how do we build more loyalty. And 77% in the survey said, that was what they were going to focus on this year, that's very easy to say, incredibly hard to do in a world, which is super competitive. And where many people just prefer to get access via platforms, because it's just so easy, so convenient. So the other thing that people are looking at is cutting costs. And trying to basically match your resources to the reset traffic levels, and then others looking at alternative channels. So we talked about some of those video based channels, which are not about referral, but it's about creating video content in Tik Tok, and, and YouTube. And then I guess the one that’s right up the top of publishers’ minds right now is WhatsApp, which traditionally has been really hard to do anything much with because it's all about organic discussion about news where it happens. But late last year, WhatsApp introduced these things called broadcast channels, and so did Instagram, actually. So you can now create, effectively broadcasts and you can take particular niches, or verticals, and start populating that within WhatsApp. And we're seeing quite big numbers. For some publishers. It's not going to rival Facebook, but I think it's, it could be quite significant. So broadly, a combination of these things. And particularly, if you're trying to reach younger audiences, you're probably going to have direct and try to use some of these new alternative platforms.

AI and journalism 

Federica: When last year’s report was published, ChatGPT had only been available to the public for a month or so. Now, the idea of generative AI is not so novel anymore. Indeed, we talk a lot about it. And news media, like other industries are figuring out how they're going to actually use it. In some cases, with experimentation first, but also shaping and beyond experimentation. What have you seen by talking to newsroom leaders, and from the survey, about how news organisations already using AI and what they plan to do more in the year ahead?

Nic: Well, I think the general consensus is that AI is going to be incredibly useful in helping journalists be more efficient, and making those workflows around journalism and much more efficient. So that's really the main focus. And by that, I mean, translation of content, transcription of interviews, copyediting, tagging of content, so it's, it's much easier to find for consumers. So a lot of this is going to be automated this year, or AI-assisted in some way, a touch of a button. And what we're seeing is, is more AI tools that are being built into content management systems, for example, to facilitate some of those workflow improvements. And in our survey, about half said that this is going to be very important to them this year. So this is really the the dominant thinking amongst publishers. Beyond that you have content creation, of course. So this is basically using AI to write stories or create summaries or generate pictures, or even create videos. And of course, that is much more risky from a trust point of view, but could also save quite a lot of money. And so we are seeing more real-life examples of that as well. So for example, in Germany, you've got a publisher that already has created an AI writer. So stories are actually written by AI and that's now accounting for about 5% of its entire publication schedule. And then there's even TV stations being launched this year, which are either completely generated by AI or partially generated by AI. So on the front end stuff in the product this year, we're going to see a lot more AI-generated content, synthetic content, mixed content. And I think that's going to raise a whole load of issues.

Perceived risks of AI 

Federica: You mentioned that some of these approaches, of course carry more risk, especially in terms of the impact on the audience. In a year where the political stakes are so high in many countries because of elections, are media leaders particularly worried about these potential risks involved using AI?

Nic: Definitely, and certainly around the content creation, or some of the people we talked to were on your way. So over half 56% of that, that was the biggest risk followed by newsgathering. And these are obviously both areas that involve the integrity of the content, so fundamentally deal with these issues of trust. Other issues such as back-end automation, or using AI for coding, making coding more efficient, or even, fact checking, these are considered to be much less controversial. And so that's why we're really seeing the effort in these back-end things. First, how can we make efficiencies? However, having said that, I think I think we're gonna see a split this year. So I think reputable news organisations that really worry about the trust of their brands are going to really focus on the back-end automation, they're not going to do so much on the product. And that's partly because we've seen the impact on trust, when AI has been used, it's not been properly labelled, or the big mistakes, it really affects the trust and the brand. On the other hand, I think you'll see a lot of other media companies that don't have much to lose, that are just basically looking for, you know, to survive effectively, really embracing a lot this AI because cost savings can be very considerable. And I think the concern about that is that that will reduce the trust in journalism as a whole. People won't necessarily make that distinction. So I think it's gonna be a really interesting year to see how publishers apply the AI.

Federica: There are many things that I think as an industry will learn from the relationship with social media. And specifically, we've seen in the past that the relationship with the social media platforms can be at times more difficult. What are we seeing in the space of AI between in terms of a relationship between news organisations and AI companies at the moment?

Nic: I mean, in many caaes, the companies are the same. So if we're talking about Google there, we've seen you and Microsoft are a new players in terms of OpenAI. But I think the publishers are really concerned about AI, and what it might do to, to their business models, and particularly around search. So we've already seen AI systems being introduced in integrated into search engines like Bing,  Microsoft Copilot, so you could ask it a question about the news. And it will surface the answer, it won't necessarily have a whole load of little links, and then you link to the link and it goes to the website, which can be monetized, that the answer is just there. And there's less reason to click on. So that's the that's the kind of big fear. And we know that news sites have been used to create the foundational models for these AI. And now, that AI is competing with which the with the news sites itself. So this has led to media companies having very robust discussions, shall we say, and in some cases, in the case of New York Times suing OpenAI and suing Microsoft, essentially claiming that they've, used the content to, to train their models, and in some cases, to produce verbatim sections of text within it. So that's what the legal case is about. And also, that it's effectively competing as a trusted news source. So we've got legal action going on the one side, and then you've got a whole load of attempts to do deals and actually try and reset the terms of trade so that you can get fair recompense for the work you've put in. And so we've had two big deals between Axel Springer, and AI companies, and also AP. And I think we're gonna see a lot more deals this year. So I think the relationship is going to be bumpy. But it's also going to be different because licensing is definitely going to be part of that discussion over the next year.

Federica: Apart from being bumpy. What's your prediction of how the full landscape is going to work out?

Nic: I mean, I think in our survey, the publishers we talked to were not very hopeful that it was going to work out very well. So I think 48% said that they didn't think that there would be much money for any publishers at the end of the day and 35% felt that if there was money, it would go to the big publishers like New York Times, or Axel Springer that had enough heft to do really big deals. So I think it's quite a lot of concern about that, I think it's really hard to know. And it's really hard to know how valuable the news content is going to be for consumers. So in some of the early research we've done, we've seen that news is not a particularly big use case, for example. So I mean, just asking for simple facts or playing around with it, or writing essays, you know, these are the kinds of things people are using AI for, rather than accessing news, people don't necessarily want more news. So the extent to which the business model is going to be undermined, we don't know. And also, the threat of legal action, I think, is also stopping some of these bots showing some of that content. So, you know, in the last few months, we've seen some of some of that shut off, because some of the AI companies are worried about the potential implications of it. An app like Artifact, for example, announced it was closing down not particularly because of this, but in recent months, you've seen, you've seen news content that it was previously summarizing, and using some of these AI tools to create different kinds of transformations, just being blocked off because they've been concerned about what the implications might be.

Plans to tackle news avoidance 

Federica: Of course, for several years, you've also tracked how audiences are disengaging from news with the work you do also on the Digital News Report. You've been talking about concept like news avoidance or news fatigue. With wars, climate change, polarising events like elections, likely to be dominating the news agenda in 2024, stories that in one way attract maybe some of the loyal audiences, but also turn off some of those people that are less inclined to engage with news more frequently or actively avoid some news some of the time, what are news organisations planning to do to deal with news avoidance, specifically?

Nic: Yeah, I mean, I think you've, you've set up the issue there, it's complicated, because you know, some things that turn people away, like politics for other people. That's what brings people to new sites. So in some senses, news publishers are hoping they're gonna have a really bumper year this year, because of all the elections around the world. On the other hand, I think the long term trends, as we've shown are, that whole sections of people are accessing a lot less than they used to. Many people, hard-to-reach audiences are becoming disengaged in different countries, a portion of those people are becoming disengaged. And as well as, that's partly to do with people, societal issues, and people becoming just disengaged with a whole load of institutions, not just news.But I think it's also a mismatch in terms of the product, in terms of what newsrooms are producing and the formats they're producing and the tone and what journalists and what audiences increasingly want in terms of, you know, accessible news that fits in with their lives in different ways. And that's really the challenge that publishers are trying to get their heads around.

And I think we have seen new types of journalism or new approaches, such as explanatory journalism, we've talked about that for a number of years. And in the survey, that's what came out on top as publishers said, we're going to invest more in that, just explaining the news better (67%), followed by what's constructive or solutions, journalism. So not just pointing out the problems, but actually offering people a sense of hope or agency, inspiring stories, you know, these kinds of approaches, particularly around something like climate change, for example, I think in the report we talked about, Bloomberg just launched a solutions podcast, but I mean, there's so many different examples of this around the world now and a lot of investment in it. And then, a little bit less interest in just creating more positive news, that seems to be a loaded term, which many journalists don't like or dumbing down and making the news more fun and entertaining. But I think a range of all of these kinds of approaches are needed. And then finally, you know, just better diversity in terms of the people you have in the newsroom, so that you're not just serving up, you know, the same old stories for the same old people, but you really are bringing a diverse set, a broader set of perspectives too.

The use of different news formats 

Federica: A bit of a provocative question, and maybe I'm mixing apples and oranges. But we know from the Digital News Report that audiences still prefer text over video in some in some contexts, although younger audiences prefer more video. But then this year, you ask publishers what they're planning to focus more of their attention on. And it seems that they say they will produce less articles and more videos. Is it because they're catching up? Or are you seeing something different in this?

Nic: It's kind of bewildering, isn't it? Basically. You know, we're all overloaded and when you ask publishers is what they're planning to do. They're going to produce more video, more podcasts, more newsletters, and very few of them say they're going to produce fewer articles. So essentially, we're just going to throw more stuff at you. And I think, you know, that's not really what audiences want. They want smarter media, they want more relevant media. And I think that is the challenge. You know, for last few years, we've just been producing more media. And that just reduces the value of the media that seems to be out there from an audience perspective quite often. So I think, you know, that's not to say that the individual initiatives, and the individual podcasts or newsletters aren't valuable, many of them are and that's why they're doing them. But much of it’s to do with the incentives. I mean, essentially, platforms and algorithms need more content, and it needs regular content. It’s not necessarily what audiences need. And again, it goes back to this mismatch. In terms of what we're providing, and what people really want.

Revenue raising strategies 

Federica: We talk a lot about these fundamental shifts in how news organisations are producing or distributing news and engaging with audiences. What shifts are you seeing in terms of raising revenue? What will publishers focus on?

Nic: Well, I think because of the problems of, you know, the end of mass referral, and the problems with volume. And we've obviously seen what’s happened to BuzzFeed News and others, I mean, the real shift is that more publishers are basically going to try and get people to pay directly, getting readers or consumers to pay directly. And we've seen that for a number of years, mixed with other models that they're trying to develop, whether that's ecommerce, licensing, from the AI platforms, of course, we've talked about that already. But essentially, ads are going to get less important. And one sign of that is the Daily Mail, the Mail Online, one of the biggest online, reached based websites in the world, has, this month, started a premium paid subscription as well. And that's just a trend we're going to see more and more of, but again, more publishers chasing the same number of people or small number of people who are prepared to pay, it's going to be really, really tough. So one thing is paid.

And then I think within that, what we're going to see this year is the growth of the bundle. So what many publishers have found that just a news product alone is not enough to keep people a lot of people say that they're gonna unsubscribe or they're not sure. So there's a huge amount of churn currently. So how can you get people to lock in more easily. So you've seen the New York Times bundling in games and product reviews and sport into a package bundle. And those bundled subscribers are much more likely to retain and monetise better over time, we see many different kinds of bundling of emerging now. So Nordic countries, it's much more about, you know, magazines, local, national podcasts as a bundle, between a number of titles, whereas the New York Times is kind of within one title. And I think this year, we may also see bundling across titles. And then the other big trend, I think, in paid is more lower price news products that don't necessarily take so much time. So that may be a podcast, it may be, you know, a newsletter. And a lot of publishers are launching those low-cost subscription products, in the hope of enticing next generation subscribers for less money.

Federica: Absolutely, I might, or might have not, just subscribed to one of those newsletter courses.

Nic: Of course, the institute podcast will remain free.

Next generation AI 

Federica: Of course, let's talk about one of a topic that I know you really care about, which is new devices and interfaces. VR headsets, smart glasses, many of these devices didn't really take off in the past, as some of the hype around them would have suggested, in terms of consumer adoption. What do you think about the Rabbit Companion, as has been defined as the computer who can talk to and will do things for you? Or Humane’s new AI lapel pin? Will they have more luck?

Nic: The interesting thing is we're starting to see new devices that they got often that smartphone killers, but they're not really, because the smartphone, of course is going to endure. But the whole idea is that we are really struggling with… the smartphone was designed as this thing that was going to be intuitive, and solve all the problems and bring everything together in one device. The reality is we have multiple passwords, we can't remember who's messaged us in which application, it seems to be getting more complicated. And we spent all our lives looking at this screen and endlessly scrolling, doom scrolling. So part of the reason we're seeing all this activity is partly AI but it's also how can you solve that problem and get the technology out of the way and that's what the Rabbit R1 is about. It's what the Humane Pin is about. It's, can you find new ways to more intuitively access information? And most of these are using essentially natural language. So you essentially talk to it. And then they have a screen or they don't have a screen. Now, people don't really want another device. So I think it's very likely the Rabbit R1 and the Humane Pin will probably fail. But I think that idea over time, there is a problem there to be solved. And I think it's gonna be really interesting to see how that works.

And we shouldn't forget, by the way, in our in our survey, people felt that the thing we should really watch is, is some of these hearable devices, because we already have headphones, we already have smart speakers in our homes that are going to be revolutionised by AI and become much more useful over the next few years. So I think that talking to a computer, when it's appropriate, getting direct responses without having to go through a whole load of passwords is going to be a huge part of the future. It's not going to replace the computer, it's not gonna replace the smartphone. And I think that's what's felt by our respondents as well. But I think it does push more in the direction of audio as becoming, you know, more important both as an input and as an output over the next five years or so.

Federica: I wanted to end with a bit of a more a more personal question, in terms of what have you found surprising in this year's findings? Of course, you spend several years doing this report and say you spend a lot of your time talking to news leaders. So you're often ahead of these trends. What surprised you?

Nic: There's lots of little things, but I think you have to really talk about the AI because it is amazing, when you use some of these cutting edge tools to see how far and fast some of the technology has moved in six months, in nine months, this is quite rare, you know, to see such big improvements, for example, you know, something like Midjourney and the quality of the images it was putting out a year ago. And the images we see today, which are photorealistic, which is very hard to tell the difference between something that's just been created by a prompt and it's actually, you know, a real photograph. And I think that, you know, that obviously raises huge issues. But from a technological point of view, that is that's absolutely fascinating. And I think the other thing is just the speed with which the industry is trying to adapt AI and find different ways of using these technologies. And that surprised me, because I thought they would be, in a way, more cautious. We've talked about the caution. But I think we're genuinely seeing now in the products, examples of AI-generated content. And I think over the next year, we're gonna see a lot more of that. Lots of uncertainty. But I think this is going to be the year when, when the template gets set for, you know, what consumer expectations are for, for how it affects trust, for how the business model has changed. So I think it's a really, really important year to help define how journalism and AI are going to sit together and work effectively together.

Federica: What a fantastic way to end the podcast. Thank you so much, Nic for joining us today.