Dutch online journalism platforms have succumbed to the echo chamber effect due to their usage of algorithmic content ‘recommender systems’ (Algorithm Watch 2019). Outlets like RTL Nieuws, Het Financieel Dagblad, NU.nl and the Dutch Broadcast Foundation (NOS) employ algorithms on their websites to advertise to readers which articles they should read next. The algorithms track a user’s reading history and interests in order to give recommendations. Blendle is an online platform that aggregates and curates content from across the internet, offering users articles on a pay-per-read model (ibid). Except for a handful of articles selected by editors, the site works by intuiting user’s reading interests from past browsing history and then curating content based on those interests.
The effects of such ‘recommender systems’ are being widely discussed, with some critics claiming that they could be ruinous for democracy (ibid). When a user only consumes content that has been carefully personalized for them, they never encounter ideas or stories that run contrary to their preset opinion, worldview, and narrative. Not only does this create a monoculture of beliefs and opinions, but it also produces an echo chamber that reverberates and amplifies ideas through circles of like-minded individuals. This process has been condemned for being conducive to the radicalization of political thought.