Quests Remakes Bluesky Attie as a Social Research Engine

On July 24, TechCrunch reported that Bluesky Attie is moving beyond custom feed creation into open-ended social research, a shift that could make the AT Protocol more searchable, more useful, and more contentious at the same time.

The new feature, called Quests, lets users ask Attie questions about news, trends, and conversations spreading across Bluesky and other apps built on the AT Protocol, according to TechCrunch. That matters because Bluesky’s growth curve has cooled. The network doubled to 40 million users last October, within a year, but now reports around 45.6 million registered accounts.

July 24: Bluesky Attie shifts from feed builder to social research assistant

Attie launched in March as a separate Bluesky product for creating custom social feeds without programming. Users could describe the kind of feed they wanted, and Attie would help build it.

Quests changes the job. Instead of only helping users design timelines, Attie can now answer questions about what people are discussing across the broader Bluesky social network, which the company calls the Atmosphere. That includes Bluesky itself plus any app that integrates with the AT Protocol.

The practical payoff is obvious. A user doesn’t have to manually dig through posts, accounts, and topic clusters to understand what’s spreading. They can ask Attie what’s gaining traction, what people are saying about a subject, or which accounts appear influential in a field or location.

That’s the core shift: Bluesky Attie is becoming a search and analysis layer for open social data, not just a feed customization tool.

March launch, July expansion: what can Quests actually answer?

Bluesky says users can ask Attie to pull information about trending topics or identify influential accounts in a particular field or location. That makes Quests closer to a research assistant than a recommendation tool.

A typical query might look like:

• Trend scan: “What topics are gaining momentum on Bluesky today?”
• Influence mapping: “Which accounts are central to this discussion?”
• Conversation research: “What are people saying about this product launch?”
• Local or field-specific search: “Who is posting about this topic in a given location or specialty?”

The value comes from compressing scattered public conversation into a digestible view. But that compression is also where the risk sits. Before using Attie’s output for publication, investment research, or public claims, users should know whether it links to original posts, how fresh its data is, and exactly which AT Protocol apps are included.

That last point is not a detail. “Across the Atmosphere” sounds broad, but the source material does not specify a full app list, data freshness rules, or citation behavior. For serious research, those gaps matter.

Jay Graber, Bluesky’s chief innovation officer and former CEO, framed Attie as a tool for cutting through noise rather than adding to it.

“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.

The AT Protocol is the reason Attie is not just another platform chatbot

The AT Protocol is the open networking system behind Bluesky. Its purpose is to let social apps interoperate instead of locking identity and data inside one company’s app.

That architecture gives Bluesky Attie a different role than an AI feature bolted onto a closed network. If public conversations exist across multiple AT Protocol apps, Attie can query a broader social layer rather than only one app’s internal feed.

This is also where Bluesky’s strategy becomes clearer. The company is not only trying to make Bluesky easier to use. It is trying to make the protocol underneath it more approachable. Graber said Bluesky wants Attie to reduce the expertise needed to use the AT Protocol and put customization tools into more users’ hands.

That fits the company’s recent AI push, including experiments around user-controlled interfaces and agent-style tools. It also connects with a wider race to make AI habitual, as XOOMAR covered in 950M Google Gemini Users Force AI Race Into a Habit War, though Bluesky’s version is tied to social data rather than a general assistant.

A fast-moving tech rumor shows where Attie could help, and where it could mislead

Take a hypothetical case: a cloud outage rumor starts spreading across Bluesky while the network has around 45.6 million registered accounts. A reporter, analyst, or incident watcher could ask Attie:

• Origin: “Who first posted about this outage?”
• Claims: “Which claims are being repeated most often?”
• Pushback: “Are credible accounts disputing the rumor?”
• Clusters: “Are developers, company accounts, or reporters saying different things?”

That could save time. It could also produce a false sense of certainty.

Attie should be treated as a starting point, not a source of truth. Users still need to click through, verify original posts, check primary company channels, and separate jokes, sarcasm, and screenshots from confirmed facts. Social platforms are messy inputs for AI systems. Rumor and evidence often look similar until a human checks the trail.

The same caution applies to AI agents in workplace and social tools more broadly. As XOOMAR noted in Jack Dorsey Buzz Hits Slack With Open AI Agent Chat, agentic interfaces are moving into communication products, but usefulness depends on how well they preserve context and accountability.

Public posts feel different when an AI can summarize them at scale

Bluesky has a cultural problem to solve. TechCrunch notes that many Bluesky users dislike AI for economic, political, and environmental reasons. Another related report from Crypto Briefing said Attie became the second-most-blocked account on Bluesky shortly after launch, with more than 125,000 blocks.

That reaction shows the tension. A post can be public and still feel different when an AI assistant can search it, summarize it, and turn it into a report.

The risks are not abstract:

• Accuracy: Social posts include jokes, partial context, brigading, and rumor.
• Consent: Users may not expect their posts to be packaged into AI-generated research.
• Abuse: Bad actors could use social analysis tools for harassment, doxxing, or surveillance-style monitoring.
• Context loss: A summary can flatten disagreement, uncertainty, or irony.

The safeguards to watch are straightforward: transparent citations, clear data boundaries, respect for deleted content, user controls, and uncertainty warnings when Attie cannot support a confident answer.

The next test: whether trust becomes a Bluesky product feature

Bluesky has room to experiment. In March, the company said it had raised $100 million in funding in a round that closed in 2025. TechCrunch reports that this gives Bluesky runway to test ideas such as Attie, potential AI services, premium subscriptions, and tools for power users.

For now, the new Attie experience is in beta, with users invited from a waitlist over the coming weeks. Graber also said Attie is being redesigned with accessibility features, a new logo, and more changes ahead.

The watch item is not whether Attie can produce clever summaries. Many AI tools can do that. The real test is whether Bluesky Attie can make open social conversations easier to research without making users feel mined, misread, or exposed. If Bluesky gets that balance right, Attie becomes a useful proof point for the AT Protocol. If it gets it wrong, the block count may become the more important metric.

Impact Analysis

• Quests could make Bluesky and AT Protocol conversations easier to search and analyze.
• The feature expands Attie from feed creation into a broader social research assistant.
• Its launch comes as Bluesky’s user growth appears to be slowing from its earlier surge.