Google adds natural language controls to Discover feed customization
Users can now use AI prompts to explicitly tell Google which topics and sources to prioritize or prune in their personalized feeds.
Google is introducing a new feature for the Google app that allows users to customize their Discover feed using natural language prompts. This update marks a shift toward explicit user control over the AI-driven content delivery system.
Under the new system, users can specify exactly which topics or sources they want to see more or less of by typing requests in their own words. The feature allows users to tell Google the exact topics or links they want to prioritize or prune. The update is currently rolling out to the Google app, with some reports identifying it as an experimental "Lab" feature. To ensure consistency, the AI is designed to remember these preferences over time to maintain the tuned feed for future visits.
The shift from implicit to explicit
Historically, Google Discover has operated as a recommendation engine relying almost entirely on implicit signals. The algorithm traditionally curated content based on a user's search history, app activity, and location data without direct input from the user on specific preferences. This update moves the platform toward a prompt-based interface, mirroring a broader industry trend where AI-driven platforms are experimenting with more direct user steering of recommendation engines.
Reducing algorithmic frustration
This transition represents a significant change in how Google handles content discovery. By moving from a purely algorithmic approach to a collaborative one, Google aims to reduce "algorithmic frustration"—the common user experience of being trapped in a content loop or seeing irrelevant topics despite changing their habits. Giving users a direct mechanism to prune or expand their interests in real-time allows for a more precise alignment between the AI's output and the user's actual intent.
The broader AI integration
This feature is part of a wider strategic push by Google to integrate generative AI into personalization across its core ecosystem, including Search and Google News. As the company moves away from static algorithms, the ability for users to "chat" with their feed settings suggests a future where personalization is a continuous conversation rather than a background process. While the core functionality is confirmed, the exact rollout timeline and the transition of this feature from an experimental Lab to a standard setting remain the primary points to watch.