Dow Jones is embracing artificial intelligence as large-language models rely on accurate news and information, says Almar Latour, chief executive of the company and publisher of The Wall Street Journal, in this interview with Lisa Granatstein, editorial director of Beet.TV, at the Cannes Lions International Festival of Creativity. Visit Beet.TV for this video (https://lnkd.in/e92Y-ZjQ) and much more! #beetcannes #canneslions2024 #newsmedia #adsales The Beet Retreat Berkshires is coming up, July 21-23! 🏞 https://lnkd.in/ewhG-SSV
Transcript
It has been a tough year for the news industry, coming on the back of many, many tough years. And it seems to be getting tougher. For Dow Jones and the Wall Street Journal, though, this has been a unique moment where we've gone on to perform probably at our very best journalistically. Our news is in demand more than ever before. Dow Jones also has other information service products that are also in high demand. And so as the news industry and the information service industry is experiencing pressure, demand for reliable information has actually gone up. And a premium for the premium price that people are willing to pay for reliable information has also gone up. And so we have been on the positive receiving end of that. If you look at the past year and you see major news events on the global stage, there's usually a ramification for the economy, for, for businesses, for certain industries. And with every peak or with every big event, you see that we get more subscriptions as a result of people needing that information at that point. And so that's a very consistent trend. So as the world is getting more complex, demand for reliable information goes up. And that's where the Wall Street Journal and Dow Jones come into play, and a few others as well of course, regrettably, that is just a handful of companies that are doing really well on the back of that. But I do think that that is a model that can be embraced more broadly in the industry. >> But as the world becomes more complicated and as we head towards the presidential election, how do you manage advertisers skittishness around being brands, safe around advertising? >> Yeah, first of all, there's so much demand for reliable information that brings in an audience, in our case exists of decision makers in business at various different levels,C suite, CEO's, but also people deeper in a business. And that's global for us. It brings in investors, people who are active on the market, but it also brings in people who have to make decisions at the kitchen table. How to navigate this very complex environment which has ramifications for your mortgage, for how you spend your personal finances. And so that's an audience that comes in as an identifiable audience. And so between the reliability and the audience, there's a lot there for advertisers that should give them comfort that they're speaking to people who really care. And we also have ad tech available tool called Safe Suite that helps brands with individual objectives around that. Amongst those three things, the inherent value of reliable information, the quality of the audience, and the tools that we have. We've seen brands that have worked with our ad tech having a much higher yield on the back of that. >> So you started talking about ad tech and technology in general. What is happening on the AI front? How is Wall Street Journal Dow Jones embracing AI? >> We have worked with AI for a very long time. A lot of our headlines are automated. We have algorithmic trading connected to algorithmic headlines. And that has existed in our world of information services for a long time. So we have deep background in that, through the journal, through Dow Jones news wires, through Factiva, which is a research database. And so there's a long history. Then you're talking about generative AI. And that, of course, has been on the rise. And there, I would say there are three big categories in which we're active. One is securing intellectual property rights, ip rights, making sure that as learning models are spreading throughout the world, that our information, which costs a lot to produce to make it reliable. It's high quality information that doesn't just get absorbed without any financial model or commercial consequences attached to that. And so we're focused on that news Corp or parent, the company has recently struck a deal with OpenAI that underscored the importance in collaborating commercially with leading players in the generative AI space. That's one, number two, a push to deploy tools internally spreads through the entire company, through every part of Dow Jones. And the Wall Street Journal will eventually have some level of generative AI involvement. For reporters, this will mean having phenomenal research tools at their disposal, having probably help with summarization, making their workflow a little bit faster. And then that capacity can be used, one, to do deeper reporting where you couldn't go before, or that capacity can be used to create more time, which you can then spend on even more high quality reporting. Then in other parts of the company, whether it's in our advertising or subscriptions, AI will play a role. And, for example, automated pricing and just any element of the business will be touched by that. So we'll see that continue. That will happen in waves. It won't happen overnight. And there is a lot of experimentation because you have to make sure that as you deploy tools, that they actually yield the results that you want. So we're all getting familiar with that. So stay tuned for how that evolves. Then the third part is perhaps the most important part, and that is meeting customer demand. And so producing, releasing new products that have perhaps as a front door or perhaps as a feature, generative AI in some form attached to it. So within Dow Jones, we just came out with our first commercial generative AI related product, which is called integrity check. It's an important part of our business, focused on risk and compliance, and it allows customers to do self help, if you will. They can do research on certain business risks themselves, whereas before you had to go interact with humans much more. And so that can now be left for a higher and deeper type of research associated with that. So you'll see a slew of products coming out in short order and then on the horizon and much more. >> So, it's early days, but what are some of the learnings or best practices that you could share with other publishers that are starting to dip their toes into the AI space? >> Let's say first look at the value of your information. Information has value you have to figure out what that is, and you have to make sure that you can secure that. I think that's a very important first step, but it starts with that awareness. Give it away, and don't give it away cheaply. If you know that it's worth more, you have to have that awareness of, this is what my archives are worth, this is what we are worth to customers. That's a start. I think experimentation, both for inside your workflow and for new products, is important. Some of these tools are very young, and so you have to figure out how they work, where they work best. In our experience, we started on smaller, more focused areas, so you have more control over what the output might be. Or you have better knowledge as to whether the outputs that come out of generative AI in relation to your content set is relevant to a customer or not. And so we take from proprietary databases at Dow Jones, proprietary data and distinctive journalism are gonna be important. They're gonna make a difference. So invest in making sure that you have proprietary data and distinctive journalism, because that will make a difference vis learning models taken from much broader data sets.To view or add a comment, sign in