Google is adding natural language customization to its Discover feed, letting users describe content preferences directly to an AI system rather than relying on algorithmic guessing. The feature, rolling out to the Google app within days, allows users to specify topics and interests through conversational input. Google's AI then adjusts the feed accordingly and retains those preferences for subsequent sessions.
The shift addresses a persistent Discover problem. Google's feed has historically surfaced irrelevant or repetitive content because its recommendations depend entirely on browsing history and implicit signals. Users have limited control over what appears. Adding explicit preference-setting through natural language removes that friction. Instead of swiping away unwanted articles or adjusting settings buried in menus, users simply tell Google what matters to them.
The feature integrates into Discover's existing interface. A three-dot menu provides access to the customization option. The AI processes descriptions and learns patterns over time, meaning repeated requests train the system on user intent more effectively than one-off interactions.
This move reflects broader industry trends around AI-powered personalization. ChatGPT and similar models proved consumers prefer conversational interfaces for complex requests. Google is applying that lesson to feed curation, where traditional settings have proven clunky. The change also positions Discover as a more competitive alternative to TikTok and other algorithmic feeds where users feel greater agency over content discovery.
However, the implementation raises questions about data handling. Google will store preference descriptions and user-submitted keywords, expanding its profile on individual interests. The company says it remembers preferences locally, but clarification on how long that data persists and whether it trains broader Google systems remains unclear.
For Google, this update serves product strategy. Discover drives engagement on the Google app, a key mobile property. Improving its feed quality and perceived control could reduce users abandoning it for social platforms or Reddit. The feature competes directly with T