Bluesky’s AI assistant ‘Attie’ is expanding from feed-building into social research, giving users a way to ask questions about what is happening across Bluesky and the wider AT Protocol ecosystem.
The new feature, called Quests, is meant to let users ask open-ended questions about trends, news, conversations and influential accounts. Instead of manually searching across posts or building custom feeds by hand, users can ask ‘Attie’ to help make sense of activity across the open social network Bluesky sometimes calls the ‘Atmosphere’.
That makes ‘Attie’ more than a novelty chatbot. It is an attempt to use AI as a navigation layer for decentralized social media, where users have more control but also more complexity to manage.
What ‘Attie’ Is Trying To Do
‘Attie’ began as an AI assistant for creating custom feeds. Bluesky’s pitch was that users should be able to describe the kind of social feed they want in natural language instead of learning how to code feed logic. That idea fits Bluesky’s broader identity: social media should be more open, portable and user-controlled.
Quests pushes that concept further. Rather than only building a feed, ‘Attie’ can help research what is happening across Bluesky and other apps that use the AT Protocol. Users might ask about trending topics, identify influential accounts in a niche, track a conversation or understand how a topic is spreading.
Bluesky’s own FAQ describes the company as a Public Benefit Corporation building open and decentralized public conversation. It also says Bluesky had more than 44 million users as of May 2026. That user base is large enough to make social research useful, but the open-protocol structure means discovery can be more complicated than on a single centralized platform.
That is the gap ‘Attie’ is trying to fill. If social media becomes more modular, users will need better tools for finding signal across many feeds, apps and communities.
Why AI Fits The AT Protocol
The AT Protocol is designed to make social networking more portable and composable. In theory, that gives users more choice over clients, feeds, moderation and identity. In practice, more choice also means more work. Users need tools to discover conversations, manage feeds and understand what is happening beyond their immediate timeline.
AI can help with that because it is good at summarizing, clustering, filtering and answering broad questions from messy data. A tool like ‘Attie’ could help users understand a fast-moving conversation without depending entirely on a platform-controlled algorithm.
That is an important distinction. On major social platforms, algorithmic feeds often decide what users see in ways that are difficult to inspect or control. Bluesky’s argument is that AI should help users customize their experience instead of serving only the platform’s engagement goals.
There is a real product opportunity here. Social media users often want to know what is happening, who is worth following and which claims are gaining attention. If ‘Attie’ can answer those questions with transparent references and user control, it could make open social networks easier for regular users.
The Misinformation Problem
Bluesky is framing ‘Attie’ partly as a way to fight noise and misinformation. That is a difficult promise. AI tools can help summarize public conversations, but they can also confidently repeat weak claims if the underlying data is messy.
The value will depend on how ‘Attie’ shows its work. If it simply gives a polished answer without explaining where it came from, users may treat the output as more authoritative than it deserves. If it surfaces sources, accounts, patterns and context, it could be a useful research companion.
This is where social AI tools need to be handled carefully. A bad summary can distort a conversation. A biased ranking can elevate the wrong accounts. A hallucinated claim can spread faster if it appears inside an assistant built for social discovery.
Bluesky users are also known for being skeptical of AI. Some object to the economics of AI training. Others worry about environmental costs, synthetic content and platform incentives. That means ‘Attie’ has to prove it is useful without making Bluesky feel less human.
The tension is similar to other AI features entering creative and social spaces. Tech Block Central recently covered how AI-generated images in galleries are forcing institutions to define AI boundaries. Bluesky now faces a different version of the same question: when does AI help people understand culture, and when does it get in the way?
Why This Could Matter For Social Search
Social search is becoming more important. Users often look to social platforms for real-time context, product opinions, local updates and niche expertise. But traditional search engines are not always good at capturing live social conversations, and social platform search can be limited.
‘Attie’ could become a bridge between search and social feeds. Instead of typing keywords and scrolling through posts, users could ask a research question and get a structured explanation of what people are discussing.
That could be useful for journalists, researchers, marketers, creators, developers and regular users trying to understand a topic. It could also help users find communities they would otherwise miss.
The open-social angle matters because the results are not limited to one closed app if more AT Protocol services grow. A research assistant that understands the wider protocol could become more valuable as the ecosystem expands.
The Revenue Question
There is also a business angle. Bluesky needs revenue to keep building the app and the AT Protocol ecosystem. AI tools could become part of that strategy if they provide enough value for power users, researchers or organizations.
That creates a delicate balance. Bluesky’s brand is built around user agency and openness. If ‘Attie’ becomes too aggressive, too promotional or too opaque, it could clash with the culture that made Bluesky appealing in the first place. If it stays clearly optional and useful, it could become a practical tool for navigating open social media.
The beta phase will matter. Bluesky needs to learn whether users actually want AI help for social research, and whether the answers are reliable enough to trust. The company also needs to make clear how data is handled and how users can opt in or avoid the feature.
For now, ‘Attie’ shows one possible future for AI in social media. Instead of generating more posts, it can help users understand the posts that already exist. That is a more defensible use case than flooding timelines with synthetic content.
Would you use an AI assistant to research conversations on Bluesky, or would you rather browse social posts manually?