Trending HCI drivers for news dissemination : a provocation for current AI news optimisation methods on social media platforms

Abstract

News dissemination practices for established news organisations have radically changed in recent years. Many news portals generate most traffic from social media referrals, that is, from readers clicking a social media post that links to an article on the portal. This shift is giving social media platforms such as Facebook a stronghold on who reads which articles. By using AI optimisation, platforms like Facebook do an excellent job in presenting readers with the most relevant ‘personalised’ articles. This ensures that readers come back to the social media platform to find the content they ‘personally’ like. This approach however spirals the readers into reinforcing existing believes and kills reading diverse sources. In this light, should there be a limit at policy level in the degree of how much social media platforms perform AI content optimisation? In other words, should policies enforce that a percentage of content on Social Media platforms is randomly selected rather than optimised for relevance and would such an approach increase the likelihood of reading news outside your bubble?