New report: Algorithms are neural, but journalists are credible
In our everyday interaction with news, it is hard to avoid algorithmic interference. The news we see on social media has been filtered by algorithms, news feeds on online news site might have been algorithmically personalised to fit our preferences and some articles might have even been written by algorithms, also known as robot journalism. That was the starting point for a new report by Arjen van Dalen from the University of Southern Denmark giving insights into what Danes think about the increasing use of algorithms in news production and distribution.
The report shows that 75 percent of Danes think that algorithms can have a role to play in the selection of news, but only in collaboration with editors and journalists. Only 11 percent believe algorithms should be solely responsible for making this decision. They also remain more critical towards the algorithmic production of news. The report, therefore, recommends that media organisations should not attempt to replace journalists with algorithms, but rather focus on making them work together in delivering the news and be selective in when to use them. This is particularly important because while algorithms are considered neutral and personally relevant in their selection, they lack in credibility, objectiveness and in ensuring a diverse and balanced selection.
The report is based on two surveys conducted by the Epinion amongst 2.435 Facebook users between the ages of 18-65. These were conducted as part of a larger research project named ‘Algorithmic Gatekeeping’. The report is directed at the industry as well as researchers as it offers recommendations on what to be aware of in relation to for example generational divides and how to market algorithms in news.
Low trust in algorithms and large generational divides
While algorithms according to the Danes are considered more neutral that does not mean that they are considered trustworthy. The report shows how Danes generally have a very low trust in algorithms used to write or select news compared to how much they generally trust both Danish news media and journalists and more general notions of algorithms, artificial intelligence (AI) and robots are considered much more trustworthy, only influencers are seen as more of equally as untrustworthy.
“It is quite interesting that there is a conflictual relationship with algorithms as part of the news, where they are at ones considered untrustworthy, but also are seen as holding potential in increasing personal relevance when they work together with journalists and editors. This report offers a great entry into initially understanding some of these divides and attitudes towards algorithms in the news, which will be helpful when we in a more qualitative manner investigate datafication of publics and how they in their everyday engagement practices with news and other platforms,” says Mette Bengtsson who is responsible for subproject three in the DataPublics project focusing on audience practices and personalised publics.
Read also: About the project "Datafied News Media – Datafied Publics?"
One of the main findings in the report is also how wide the generational divide proved to be. There where major differences between the generation that grew up with social media as they were much less hesitant towards the use of algorithms. An observation that will also be important moving forward with research on the relations between audiences and algorithms.
The report is in Danish and is available online. The project is funded by the Independent Research Fund Denmark and the Carlsberg Foundation and is conducted as part of the work in the Media Research and Innovation centre (MFI) at University of Southern Denmark.