Facebook is giving users a more direct way to influence what appears in Feed and Reels. Instead of relying only on likes, comments, viewing behavior and other passive signals, people can now tell the platform what they want to see in their own words.
Meta calls the feature Your algorithm. It is a small interface change with a much bigger implication: recommendation systems are becoming something users can actively negotiate with.
Facebook Makes Recommendation Feedback Conversational
Your algorithm lives inside the three-dot menu and lets people add and manage the topics they want to see in Feed and Reels. Meta says users can describe their interests directly, rather than waiting for the ranking system to infer them from behavior.
The update also expands the feedback that follows the “Interested” and “Not interested” options. Facebook may ask follow-up questions, allowing people to explain what they liked or disliked. Users can then describe what they want more or less of, with recommendations updating in real time, including within the same session.
That matters because most feed controls are still built around blunt reactions. A like can signal approval, but it cannot easily explain that someone enjoyed one part of a post, wants more content from a particular community, or is temporarily interested in a topic without wanting it to dominate their feed.
Natural language gives Facebook more context to work with. Meta says large language models can interpret interests, intent and context alongside the thousands of behavioral signals already used by its ranking systems.
The Feed Is Becoming Something Users Can Tune
The obvious benefit is relevance. The more important shift is control. Facebook is presenting the algorithm less as an invisible force that users must adapt to and more as a system they can actively adjust.
That does not mean people are taking over the ranking system. Meta still decides how those instructions interact with engagement data and other signals. But the interface changes the relationship. Users are no longer limited to accepting recommendations or training them indirectly through repeated behavior.
For Facebook, the move also offers a way to make personalization feel more responsive without abandoning automated ranking. The platform can keep its recommendation machinery while giving people a clearer sense that their preferences matter.
The strategic consequence is straightforward: the next generation of feeds may be shaped as much by conversation as by consumption. Facebook is not simply making its algorithm smarter. It is making the algorithm something users can talk to.