The big picture: The addition of dislikes and social neighborhood mapping underscores Bluesky's continuing effort to balance personalization, community engagement, and moderation. As user activity increases and critical feedback influences product development, the company's success may depend on how effectively these systems deliver a sense of meaningful control to its users.
Bluesky recently reached a major milestone, announcing that it now serves 40 million users worldwide. Alongside this achievement, the platform unveiled a series of updates aimed at shaping its digital environment – most notably, the introduction of a Dislike feature designed to improve content personalization across its Discover feed and other areas.
At its core, the new Dislike option functions as a private user tool, allowing people to indicate the types of content they prefer to see less of. Bluesky's engineering team will use these signals to fine-tune the ranking of posts and replies, tailoring recommendations to individual activity and interests.
While the platform's technical roadmap already emphasizes decentralized moderation – enabling users to create and apply their own blocking, muting, and filtering rules – these updates represent a deeper investment in automated personalization.
We really are anti torment nexus. You can't get rid of all the torment but, you know, you gotta put in some hours to fight it
– Paul Frazee (@pfrazee.com) October 31, 2025 at 3:12 PM
[image or embed]
Bluesky says that dislikes aren't just another engagement metric but a key part of a broader ranking overhaul. As users indicate their content preferences, the platform's new "social neighborhood" mapping system will use this feedback to prioritize posts and replies from accounts more relevant to each user's interests and social circles.
On rival platforms like Threads, algorithmic feed ranking has often led to user frustration, with unrelated or contextless content surfacing in timelines, a problem Bluesky hopes to solve through more granular controls and localized conversation mapping.
Beyond dislikes, Bluesky's latest slate of updates focuses on conversation quality and moderation. A new reply-detection model is being used to identify and downrank responses deemed toxic, spammy, off-topic, or posted in bad faith. Instead of removing problematic posts outright, Bluesky's approach is to make them less visible by adjusting their placement in threads, search results, and notifications.
The platform is also testing an updated Reply button: users are now taken to the full thread before composing a response, a change intended to reduce fragmented or redundant replies.
Amid an ongoing debate over platform-level moderation, Bluesky has faced criticism for its restrained approach to banning accounts. Some users have expressed frustration with its reluctance to remove profiles accused of violating community rules directly.
In contrast to centralized content policing, Bluesky's leadership argues that user-directed tools such as moderation lists, muted word filters, and customizable notification controls are more effective in fostering a balanced, self-managed experience. Still, the company acknowledges that such strategies may reinforce "filter bubbles," limiting exposure to dissenting viewpoints and reducing opportunities for constructive dialogue.