The REM Lab is researching the use of AI in the media industry to develop transparent and explainable AI systems, investigate the value of actions and predictions across media platforms, and create methods for automatically generating content to support independent and transparent news and media organizations in the Netherlands. REM Lab is a collaboration between DPG Media, the Jheronimus Academy of Data Science, and the University of Amsterdam.

‍

AI has the potential to contribute to the functioning of the media along the entire production chain (sourcing, production, distribution, engagement), and even fundamentally change some aspects of the media as we know it. The goal of this lab is to explore the potential of AI-driven solutions for the media, their professional role, opportunities for new business models, and new ways of engaging with users while taking into account professional values and the emerging regulatory framework for AI. Furthermore, the REM lab aims to explore ways in which media platforms can develop and grow without compromising on responsibility, yet still allowing the overall business to prosper.

‍

Our vision is that independent and open, technology-driven news and media organizations have a key role to play in any democratic society that is increasingly dominated by a small number of global players. Given (1) that AI will play a leading role in news and media and (2) that the EU is now publishing AI guidelines, the time is right to accelerate trustworthy AI technology for the entire news production chain. In our lab we will work with DPG Media, which plays a crucial role in Dutch society: reaching 92% of the Dutch population, it is the largest media company in the country. AI solutions in the news production chain at DPG Media impact all of the Netherlands: the lab will help develop and analyze AI innovations to ensure that this impact is positive. In particular, AI at DPG Media plays an important role in informing the public. Alongside their journalists, AI will enable DPG Media to offer a much more diverse pallet of information to their readers. Furthermore, AI will enable DPG Media to make the online advertising world more balanced, where DPG Media is in control of its own data processing and algorithmic processing, thus creating a much more sustainable situation for advertisers in the Dutch media landscape.

‍

Our vision of the importance of developing AI technology to sustain an independent and transparent news and media landscape in the Netherlands translates into a set of key questions that we aim to investigate. With a special focus on identifying the concrete information needs of (professional) users and designing and testing explanations that are actionable and trustworthy, we will investigate normative aspects of transparency and explainability of AI-based systems for news. A second challenge is to explain (news) recommendations to multiple stakeholders along the news production pipeline in an actionable way. To enable the applicability of AI-based solutions, we will investigate the value of actions and predictions across media platforms, both from a modelling and an evaluation point of view. And a final challenge we will address concerns reliable methods for automatically generating content.