Facebook has launched a new artificial intelligence (AI) tool designed to give users greater control over their algorithms. The company, led by Mark Zuckerberg, introduced a feature called “Your Algorithm,” which allows users to customize the topics and content they see in their Facebook Feed and Reels. Users can share feedback and content preferences in their own words. This new tool leverages large language models (LLMs). The social media giant has updated the three-dot menu on Facebook to include this new customization tool.
Meta Launches “Your Algorithm” Feature
In a blog post on Wednesday, Meta announced the launch of “Your Algorithm,” a feature that influences the ranking and recommendation of content users see on Facebook. Through this feature, users can directly add and manage the topics they wish to see in their Feed and Reels. This dedicated new section can be accessed via the three-dot menu found on Facebook posts.
Meta states: “You can now fine-tune your experience and engage directly with the algorithms that rank and recommend the content you see on Facebook—meaning you’ll receive recommendations that are more relevant, nuanced, and aligned with your personal preferences.” Meta also notes that users can think of this as a “dedicated space to tell us exactly what you like.”
Facebook has also expanded the feedback options for posts in the News Feed and Reels. With this update, users can provide feedback on posts by selecting either “Interested” or “Not interested.” Meta may follow up with additional questions to better understand what you like or dislike. This allows users to directly inform the platform about their topical preferences and the types of content they would like to see more or less of.
Meta assures that recommended content will update in real time as users input new preferences into the system. Previously, the company relied solely on a user’s past likes, comments, and viewing history to determine what content to display.
The social media giant stated that its “Your Algorithm” tool uses large language models (LLMs) to integrate the interests, context, and intent underlying natural language.