The discussions around the experiment focus on early stage testing and the core approach of gathering user signals to adjust recommendations.

The concept of social proximity is part of the underlying framework guiding how content is surfaced and filtered.

Bluesky’s culture emphasizes blocking and selective exposure, which can create echo chambers if recommendations don’t shift significantly with those signals.

User experiences with Discover are mixed: some find parts engaging, but many feel overwhelmed or monotonous, with top-of-feed content dominating.

If the dislike feature works as intended, Discover could become more useful by filtering out broad undesired categories; if not, blocking remains the default method for curating content.

Bluesky is testing a dislike button to signal posts users want to see less of in their Discover feed, aiming to improve conversations and overall discovery quality.

A robust dislike mechanism could transform Bluesky’s Discover tab from a cluttered, repetitive feed into a more diverse, controllable discovery experience.

The dislike signal would provide feedback on content to deprioritize, potentially refreshing the Discover feed by reducing low-value or repetitive posts.