iWorld
Your feed, your rules: Meta gives Facebook users a say in their feeds
Meta’s new Your Algorithm feature lets users describe their interests and refine recommendations in real time
NEW DELHI: Facebook is giving users more direct control over their content recommendations, introducing AI-powered tools that let them describe their interests in their own words and fine-tune what appears in Feed and Reels.
The new feature, called Your Algorithm, allows users to add and manage topics they want to see. Available through the three-dot menu, the tool is designed to make recommendations more relevant by letting users communicate their preferences directly rather than relying solely on their activity on the platform.
Meta, Facebook’s parent company, said the feature would allow users to adjust their content preferences as their interests change, with recommendation systems responding in real time.
The company is also expanding feedback options for individual posts. After selecting ‘Interested’ or ‘Not interested’, users may be asked follow-up questions to explain what they liked or disliked. They can then describe the type of content they want to see more or less of, helping Facebook refine recommendations during the same session.
The update marks a shift towards more interactive content ranking, with large language models (LLMs) helping Facebook interpret users’ preferences expressed in natural language.
Facebook’s existing recommendation systems use multiple signals, including likes, comments and viewing behaviour, to infer what users find relevant. Meta said LLMs would complement these signals by helping the systems understand the context, intent and nuances behind users’ stated interests.
By combining freeform text with existing engagement data, the company aims to make Feed and Reels recommendations more responsive to users’ immediate preferences.
Meta has also updated the three-dot menu on posts to make content customisation tools easier to access. The changes give users additional ways to influence what they see, while retaining the platform’s existing recommendation signals.
The rollout puts greater emphasis on conversational feedback as a way to shape social media feeds, allowing users to communicate their preferences directly instead of leaving algorithms to infer them entirely from their online behaviour.
