YouTube will let users create their own recommendation algorithm with AI
YouTube is introducing AI-powered custom feeds that allow users to create personalized recommendation streams using natural language prompts. The feature uses Google's Gemini AI model to generate feeds based on detailed descriptions of interests, moods, viewing habits, or content preferences. Users can request anything from podcasts for a daily commute to relaxing videos for unwinding after work. Once created, custom feeds are pinned as separate tabs at the top of the YouTube homepage for easy access. The feature reflects a broader industry trend toward user-controlled recommendation algorithms, following similar initiatives from Bluesky, Threads, X, Instagram, and Spotify. YouTube believes custom feeds will help people discover relevant content among the platform's more than 20 billion videos. Importantly, these feeds will supplement rather than replace YouTube's main recommendation feed. Support for multiple custom feeds will begin rolling out on web and mobile next month.