What happened at the World News Media Congress 24?
Hi,
My name is Sara Forni, AI Product Manager at Atex, and this is MyType, a newsletter dedicated to journalism, innovation, and artificial intelligence.
If this mail was forwarded to you and you would also like to receive future releases, you can subscribe to MyType here. You can now also follow MyType on Linkedin, Instagram and Twitter.
Now, let's get started!
I know, I know… we haven't spoken in almost a month. But it has been a very busy month, full of trips, exciting workshops, and very interesting new people.
By the way, last Thursday marked the conclusion of the World News Media Congress organized by WAN-IFRA, which this year was held in Copenhagen. The theme of the 75th edition of the congress was "Shaping the Future of News Media in the AI Era" and expectations were more than met.
There were dozens of panels and talks dedicated to how Artificial Intelligence will change and has already changed the world of journalism. Among the speakers were notable names, including Alex Hardiman, CPO of The New York Times, and Anna Bateson, CEO of the Guardian Media Group. We also took the stage, but we'll talk about that in a moment.
In the past few days, I have had the time to absorb the various ideas and organize my thoughts. These thoughts will later translate into my work. In general, my first impression is that we are all more or less at the same point. Don't get me wrong, there are companies and newsrooms that have used GenAI in new and innovative ways, but generally, the impression is that by May 2024, some functionalities have become the solid foundation of innovation. Summarization, title and lead suggestions, translation (so much so that during the congress it was possible to follow the conferences with automatic transcription and translation in one's own language), content personalization, and SEO optimization are all activities that many have integrated into their systems and that many more have at least tested. It would suffice to open ChatGPT to try one of these functions.
From here, however, emerges a deeper reflection shared by many experts: the importance of customization as a key element in the success of an AI-based project. A newsroom that decides to use a language model to obtain a list of suggested titles for its content cannot be satisfied with a simple drop-down menu with a list of pre-set tones of voice (informal, lively, brilliant, serious), but must take the next step and try to replicate a writing style that reflects that of the publication. This is a simple but useful example to explain the concept od the real difference that will, from now on, be made by content personalization, with an initial phase of fine-tuning and continuous optimization of results based on journalists' feedback.
GenAI is a UX Revolution
Another important theme is integration, which is essential for making the use of AI smooth in various parts of the daily workflow. Understanding how to integrate different AI technologies in the same place and, most importantly, making them pleasant to use will be a decisive and differentiating element. While today the most well-known and widely used Large Language Models (LLMs) provide (more or less) similar performance, the user experience (UX) of digital products will play a fundamental role in determining the performance of these tools and in redesigning editorial products.
Six steps for AI adoption in the newsroom
As I mentioned earlier, at Atex we also gave a speech at the congress. The subtitle of our session was "The Evolution of Newswork and Ethics in a Generative Information Ecosystem" and we invited Hannes Cools, Assistant Professor of 'Human Factor in New Technologies' at the University of Amsterdam, to speak with us. I had previously interviewed him for a newsletter, which you can find here.
After his presentation, I discussed six steps for AI adoption in newsrooms. For about a year, we have been working with a long-standing Atex client, to integrate some AI features into their CMS, that is also one of our product. We have translated our job in six steps for AI adoption in newsrooms. These points are not strict rules because every case is different, but they help us to manage our AI adoption process.
The first step starts from the problems of the newsrooms. Only when we have identified problems, we can start thinking at the right AI solutions. With the solution in our hand, or better, the idea of the solution, we will start working closely to the newsrooms for the co-design of the AI product. In this phase, we fine-tune the outputs and try to identify the goals to be achieved.
Third, based on our experience, we have decided not to rely on a single AI model but to be flexible in our adoption. For this reason, the AI team constantly evaluates which LLMs work best for a particular request. The fourth point is about the concrete implementation of the AI solution, which we can do natively with Atex products or through APIs.
Fifth, we are a tech company and therefore believe in the value of data. For this reason, we provide to every clients a dashboard to monitor the AI consumption of various tools. This way, they can also understand when, in some circumstances, it might be time to take a step back. Sixth, of course, integrating AI into a system is a continuous cycle. This iterative process helps in refining the AI applications to better meet the evolving needs of the newsroom and ensuring they deliver the intended benefits over time.
Finally, a story that struck me
Among the various sessions I attended, I was particularly struck by the presentation from Onet.pl, online portal in Poland, with 25 years of experience in the digital publishing and internet industry. Like many digital media outlets and beyond, Onet has had to rethink how to engage users and "do" journalism in recent years.
Founded in 1999, Onet continued on its path like most digital media until 2018. At some point, however, they realized that something needed to change, and it was necessary to engage users (who in their case were a vast audience with very different interests) in a targeted way. Thus, from 2021, they began a journey of content personalization for audience segments. With the boom of Generative AI, since 2024, they have practiced what they call 'one-to-one personalization,' meaning the ability to customize the website's homepage for each individual user. So, a young teenager interested in sports will see the news arranged in one way, while a 60-year-old woman will see different news, with different headlines.
Now, 18 million Poles visiting Onet every month enjoy personalized and engaging home pages.
Big opportunity!
Have all these news inspired you? Would you like to integrate AI into your editorial workflow as well?
Contact our AI team or send me an email (sforni@atex.com) we can do it together!
There would be no 'digital joynalism' without sharing. If you like this newsletter, you could share it with colleagues, friends and family :)
Have a good weekend,
Sara