YouTube is giving users a major power-up, introducing a new "Custom Feeds" feature that allows for direct manipulation of the video algorithm. This move marks a significant shift, empowering viewers to shape their homepage content beyond simple likes or dislikes.
The new feature, announced at YouTube's Made on YouTube event, lets users type in natural language requests to curate their ideal video feeds. Imagine asking for "rainy Shanghai vibes with cafe jazz, street food, and vlogs," and YouTube's AI generating just that. Previously, influencing the algorithm relied on less direct methods like reacting to videos or blocking creators.
Emily Moxley, YouTube's vice president of product management, highlighted the demand for greater user control. "Users are always telling us that they want more control," she stated. "They want more ability to steer their algorithm. With custom feeds, we're giving them a natural-language way that they can do that."
Testing has shown the feature's effectiveness. One user requested "gooey macaroni and cheese recipe videos, Caleb Hearon clip compilations, and crystal singing bowl videos under 15 minutes," and received a perfectly tailored mix. YouTube even offers suggestions to refine prompts, such as filtering by upload date or excluding commentary, adding another layer of customization.
This feature isn't just about simple topic selections; it excels at blending multiple preferences, offering dynamic responses to complex requests that go beyond standard search. It's a low-stakes environment designed for experimentation, ensuring users can refine their experience without fear of "messing up" their main feed, which remains accessible via an "All" button.
Custom Feeds arrive as part of a broader trend across tech platforms. Meta's Instagram introduced a "Your Algorithm" feature, and Spotify launched "Prompted Playlists," all aimed at giving users more agency over their personalized digital experiences through AI-powered tools.
