X Admits Algorithm Issue Made Site a 'Battleground'
X's head of product, Nikita Bier, admitted on Monday that the platform's algorithm failed to surface posts from followed users. A planned tweak aims to boost visibility for posts to 'mutuals,' potentially enhancing community.
X, formerly Twitter, acknowledged on Monday that its recommendation algorithm had a significant flaw. Nikita Bier, the company's head of product, stated in a post that the algorithm was "missing" data related to surfacing posts from people users actually follow.
Bier announced that a forthcoming adjustment will "boost visibility of your posts to your mutuals," aiming to foster a greater sense of community on the platform. This change is intended to reduce the prominence of random arguments and divisive content that had become prevalent.
He did not, however, explain how the critical data point initially went missing from the algorithm's processing. This admission follows recent statements from Bier regarding content policy changes.
Last week, Bier also commented that users posting original content would "climb faster," after admitting that top accounts had been sharing "stolen" recycled content. The new algorithm tweak may address the platform's previous tendency to amplify certain types of content over original contributions.