Technology news and discussion on TalkFuse.
X just tweaked its algorithm to make it more friendly, less battleground
X has adjusted its recommendation algorithm to give greater visibility to posts from "mutuals" — accounts that a user follows and who follow them back — in an effort to make the platform feel less combative. Head of product Nikita Bier said the previous algorithm had been missing this data, which caused friends to appear less often in reply sections and made interactions feel like "a battleground" full of strangers. The change is intended to make X feel more like a community rather than a stream of disconnected voices, and to help users cluster around shared interests, a feature Bier said many had requested.
The update is part of a broader push by X to make the platform more appealing to creators, following earlier changes this year to how it compensates accounts for original content and the recent launch of a built-in video editor. It also echoes moves by Meta's Threads, which last month introduced a "Your Algo" tool letting users control their feeds privately, as it seeks to differentiate itself from X; Threads has since reached 500 million monthly active users.
In short
- X now boosts posts from mutual followers in replies and feeds
- Aim is to reduce hostility and foster interest-based communities
- Follows similar community-focused algorithm changes at Meta's Threads
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Prioritising mutual connections reflects users' genuine preferences for connecting with trusted contacts rather than strangers in conflict. The previous algorithm's emphasis on engagement inadvertently amplified combative dynamics that discouraged authentic community formation. By aligning the algorithm with users' actual relationship graphs, X makes the platform feel less like a battleground and more like the social space users actually want, which research and user feedback consistently demonstrate they prefer.
The case against
Algorithms that narrow feeds to mutual followers risk creating insular echo chambers where users encounter only viewpoints their existing circles already share. This approach contradicts the value of social platforms as spaces for encountering diverse perspectives that broaden understanding and challenge assumptions. Greater algorithmic personalisation can inadvertently trap users in filter bubbles that reinforce existing beliefs, potentially fragmenting public discourse and reducing exposure to the challenging-but-important information democratic citizens need to encounter.
Explained simply
X, formerly called Twitter, is a social media platform where users post messages and respond to one another. Its recommendation system helps decide which posts people see, including in reply sections; “mutuals” are accounts that follow each other.
The platform is owned by Elon Musk, and Nikita Bier is its head of product. X competes with services such as Meta’s Threads, where people also share short posts and build followings.
How posts are recommended can shape whether users see familiar voices or encounter people they do not follow. The changes sit within X’s wider efforts to attract creators and give users more reasons to spend time on the platform.