On October 13, 2026, Meta officially announced the rollout of a new artificial intelligence-powered toolkit called “Your algorithm” across Facebook Feed and the short-form video format Reels. This feature allows users to directly express what content they want to see or limit using everyday natural language, rather than relying entirely on previous passive behavioral guesswork.
The new tool is integrated directly into the three-dot menu on each post for easy user access. Following its initial debut on Instagram Reels in December 2025 and expansion to discovery sections in June 2026, bringing this feature to Facebook marks a step toward completing the integration of language models into the content distribution engines across Meta.
Shifting from Behavioral Inference to Direct Conversation
Alongside a dedicated topic management interface, Meta added an interactive feedback mechanism directly on posts. When users tap the Interested or Not Interested buttons, the system can prompt follow-up questions allowing users to explain their specific reasoning in natural language. These responses are processed by the system to adjust the content stream within the same session, providing an immediate personalized experience.
From a technical infrastructure perspective, this represents a significant transition for feed-ranking algorithms. Previously, Facebook’s systems primarily inferred preferences based on thousands of implicit signals such as likes, comments, shares, or dwell time. Introducing large language models into the ranking pipeline allows machines to comprehend the context, emotional nuances, and true intent behind user descriptions, resolving misaligned recommendations caused by misinterpreting passive scrolling behavior.
Impact on User Experience and Algorithmic Transparency
Meta’s move reflects an effort to return feed control to users amid widespread criticism that social networks trap viewers inside closed information bubbles imposed by automated algorithms. By allowing users to intervene directly using language, the boundary between content consumers and the distribution mechanism becomes more transparent and interactive.
Nonetheless, the real-world effectiveness of this tool still depends on how actively user communities continually refine their personal preferences. An algorithm capable of direct dialogue is undoubtedly an intriguing technological advancement, but whether it will be sufficient to retain users amid fierce competition from rival short-video platforms remains a question that only time will answer.