Re:plus constitutes a specific method for journalists, supported by innovative AI, to analyze debates on their news site. This should allow them to provide an accurate response to identified fallacies and conspiracy theories prevalent among citizens, through disagreement checks and other forms of journalism from an autonomous editorial mission.
An important condition for a healthy debate is that there is at least a “shared reality” on which people can differ in opinion. Disinformation complicates this and promotes polarization. When discussing the role of the media in combating disinformation, people often expect just one thing: fact-checking.
However, it is not always clear what is true or false. In the case of a manipulated image, for example, it is usually clear that it is disinformation, but much disinformation is in the gray area between factual reporting and opinion. Who determines what is true or false? It is not up to the government, the media, or some independent body to determine what is true or false. It is the subject of debate.
The better citizens understand the perspectives and arguments of other citizens, the less susceptible they will be to fallacies, conspiracy theories, and inaccurate information in their discussions with fellow citizens, both online and offline. The development of Re:plus will also be accompanied by the development of methodologies and knowledge for journalists to work with. All of this will be monitored through output-oriented academic research.
Introducing Re:plus, an AI-powered tool that helps media users share their opinions through online debate, discover those of others, and shape and adjust their own views, resulting in less polarizing and more constructive societal debate. Re:plus builds on Rhetoric 4, a prototype developed within the eponymous ICON project to facilitate online debate.
The current Rhetoric toolset includes a simple flow where readers can leave an opinion and their argument. Moderation is supported by AI. Readers can view arguments from other citizens. The newsroom can consult the arguments of citizens in a purely functional overview. This flow works well in proof of concept and beta-testing, and is therefore a good basis to build on. However, we note that the application can still become much more powerful and innovative, and can therefore be rolled out more widely by linking it to misinformation, both for readers and the newsroom.
Currently under development, the project is still ongoing.