Bluesky wants its users to do more than scroll. The company just gave its AI assistant Attie a major new job. No longer content with helping craft custom feeds, Attie now tackles open-ended questions about what spreads across the network and its connected apps. The move arrives at a delicate moment.
Bluesky added 5.6 million accounts since last October. That pace lags the explosive doubling to 40 million users the year before. Executives see tools like Attie as one path forward. They hope it keeps people engaged. They also hope it creates paths to revenue. TechCrunch first reported the expansion.
The new capability goes by the name Quests. Users type a query. Attie responds with information pulled from Bluesky itself and any service built on the AT Protocol. Call the whole thing the Atmosphere. One prompt might surface trending conversations. Another could map influential voices in climate science or local politics. A third might assemble a daily briefing tailored to your interests.
Jay Graber laid out the vision. She stepped down as Bluesky CEO in March. She now serves as chief innovation officer. “Attie’s AI is a tool that helps you make sense of the internet, helping you fight back against the proliferation of misinformation and noise by giving you the tools to find the truth yourself,” Graber wrote. She added a broader point. “Our bet is that tools that help people make sense of the world will prove to be more useful than those that fill it with confusion.”
Her words carry weight. Bluesky has long positioned itself as an alternative to centralized platforms plagued by algorithmic bias and echo chambers. Yet many users on the service remain skeptical of AI. They cite training costs, energy demands and fears of job displacement. The company must thread that needle. It needs to innovate without alienating its base.
Attie first appeared in March. Back then the focus sat on feed creation. Users described the kind of content they wanted. The AI turned those instructions into working algorithms. No coding required. Early testers called the experience refreshing. Some built feeds around niche academic topics. Others filtered for calm political discussion. The promise always pointed toward something larger.
Quests takes that promise further. It turns Attie into a research partner. Think of it as a search engine tuned exclusively to decentralized social data. The difference matters. Traditional search crawls the public web but often misses real-time chatter inside apps. Attie operates inside the Atmosphere. It sees posts, replies, reposts and threaded discussions as they unfold. That gives it an edge on timeliness.
Bluesky raised $100 million last year. The round closed in 2025. That capital buys breathing room. Toni Schneider, formerly of Automattic, now steers the company as CEO. Under his watch the team plans further Attie updates. A redesigned interface will add accessibility options. A new logo is on the way. More capabilities will roll out over coming months. For the moment Quests sits in beta. Selected users receive invitations from a waitlist.
The timing feels strategic. Other social platforms experiment with AI summaries and chatbots. Meta, X and others integrate large language models into recommendation systems and search. Bluesky chooses a narrower lane. It bets on transparency and user control. Because the AT Protocol remains open, anyone can build on the same data. Attie simply makes that data easier to query.
Critics still raise questions. Can an AI truly cut through misinformation without introducing its own biases? Training data reflects the network’s demographics. Early Bluesky users skewed toward journalists, activists and tech workers. Their conversations shape what Attie learns. The company says it designs the system to cite sources and encourage verification. Results include links back to original posts. Users can dig deeper themselves.
Recent coverage echoes the excitement and the caution. AI Chat Daily noted the same slowdown in user growth and highlighted Quests’ ability to query trends across the protocol. The piece appeared hours after the TechCrunch report. It added fresh context on potential applications for journalists and researchers tracking fast-moving events.
Bluesky’s own announcement on the platform generated immediate reaction. Jay Graber posted directly. She described Quests as an interactive research and exploration tool. Users responded with a mix of enthusiasm and familiar AI skepticism. One thread asked whether Attie would eventually power paid premium features. The company has floated that idea before. Nothing is decided.
Look closer at the architecture. Attie runs on models from Anthropic. That partnership dates to the initial launch. Bluesky selected Claude for its balance of capability and safety focus. The Quests feature likely layers additional retrieval mechanisms on top. It pulls live data through the AT Protocol’s open APIs. The combination lets the assistant answer questions no static search index could handle.
Industry watchers see parallels with earlier experiments. Perplexity built a search engine on top of large language models. It cites sources and reduces hallucination through real-time retrieval. Attie follows a similar logic but stays inside one protocol family. That narrower scope could produce higher accuracy. It could also limit its appeal to users who live entirely inside Bluesky.
Yet the bet feels characteristic of the company. Bluesky grew by offering an antidote to X’s chaos. It emphasized moderation, user choice and open standards. Attie extends that philosophy into the AI era. Instead of letting an algorithm decide what you see, it hands you the controls. Instead of feeding you answers, it equips you to investigate.
Success is far from guaranteed. User growth must stabilize. Retention numbers matter more than registered accounts. If Quests delivers genuine insight without adding noise, it could become a daily habit. If it feels like another chatbot spitting generic summaries, users will ignore it. The beta period will prove decisive.
Graber sounded optimistic in her post. She believes tools that lower the barrier to understanding the AT Protocol will attract developers and power users. She wants Attie to help anyone customize their experience. That ambition stretches beyond simple feed tweaking. It points toward a future where social media feels less like a consumption machine and more like a collaborative research environment.
Plenty of obstacles remain. Compute costs for AI inference add up. Moderation of AI-generated content presents new headaches. Regulatory scrutiny of both social networks and large language models continues to mount. Bluesky must manage all of it while staying true to its decentralized roots.
Still, the direction is clear. The company refuses to treat AI as an afterthought or a pure monetization play. It treats the technology as a way to fulfill its original promise. Make social media useful again. Give people agency. Cut through the noise. Attie’s expansion represents one concrete step down that path. Whether it leads somewhere meaningful depends on what users actually do with it.
One thing seems certain. The conversation about AI in social media just gained a new chapter. Bluesky wrote it in public, on its own protocol, with its own assistant. Observers will watch closely to see how the story unfolds.
WebProNews is an iEntry Publication