**MENLO PARK, Calif.** — Threads, Meta Platforms' rapidly growing social networking platform, officially launched its "Dear Algo" feature on February 11, 2026, granting users unprecedented direct control over their algorithmic feeds. This innovative, AI-powered tool allows individuals to temporarily shape the content they see, a move that coincides with Meta's strategic push into open-source artificial intelligence, aiming to democratize advanced AI models while refining its competitive edge.
The "Dear Algo" feature operates through a simple yet public mechanism: users compose a Threads post beginning with "Dear Algo," followed by a request to see more or less of specific topics. For instance, a user might type, "Dear Algo, show me more posts about podcasts" or "Dear Algo, less content about that TV show I haven't caught up on." The platform's AI system then adjusts the user's feed according to this command for a period of three days. Connor Hayes, Meta's head of Threads, emphasized the importance of personalization, stating, "The best algorithms are personalized," and acknowledged that the feature was inspired by a viral trend of users making similar requests to the algorithm. Notably, these requests are public posts, fostering a community aspect where users can repost others' "Dear Algo" prompts to apply the same preferences to their own feeds.
This launch arrives amidst significant momentum for Threads, which has rapidly expanded its footprint in the social media landscape. The app now boasts over 400 million monthly active users. A report from analytics firm Similarweb in early 2026 indicated that Threads had surpassed X in daily mobile usage, with approximately 141.5 million daily active users on its mobile app as of January 7, 2026, compared to X's estimated 125 million. Time spent on Threads also saw a notable increase, jumping 20% in the fourth quarter of 2025.
Beyond enhancing user experience on Threads, the "Dear Algo" feature reflects Meta's broader and aggressive investment in artificial intelligence. CEO Mark Zuckerberg has articulated a vision of merging large language models with the sophisticated recommendation systems that power Facebook, Instagram, Threads, and Meta's advertising ecosystem. Zuckerberg has previously stated that the company's current recommendation systems, while driving "meaningful growth," are "primitive compared to what will be possible soon." This ambitious direction is underscored by the company's plans to release open-source versions of its next-generation frontier AI models, codenamed "Avocado" (a large language model) and "Mango" (a multimedia file generator), both slated for release this year.
Leading this charge is Alexandr Wang, who joined Meta after the company's $14.3 billion investment in his AI startup Scale AI last year. Under Wang's leadership, Meta is pursuing a "hybrid" open-source strategy for its advanced AI models. While committed to open access for many of its developments, particularly to democratize U.S.-built AI options for developers globally, Meta intends to keep certain components or more advanced versions proprietary. This balanced approach aims to mitigate potential safety risks and maintain competitive advantages in a rapidly evolving AI industry, ensuring that Meta can safeguard its most powerful innovations.
The integration of "Dear Algo" on Threads and Meta's evolving open-source AI strategy signal a future where users may wield greater, more intuitive control over their digital environments, moving beyond traditional preference settings. This shift positions Meta as a key player in shaping both the immediate user experience on social platforms and the broader development trajectory of artificial intelligence. The success of this hybrid model, balancing open access with proprietary elements, will be crucial in determining Meta's standing against rivals and its ability to translate cutting-edge AI research into widely adopted consumer products.