signulll: Reconstructing Personalized Information Streams with 22 AI Agents
signulll has revealed that its new company is building "Facebook News Feed 2.0," which continuously generates content through 22 AI agents, embedding a highly personalized information stream directly into mobile home screen components, creating a real-time updated "For You" information interface.
The product is centered around two desktop components: a medium-sized component that provides the most relevant information at the moment, and a large-sized component that carries the complete information stream. This approach contrasts with Meta's early news feed mechanism, as the content is no longer sourced from social relationships or creators but is automatically generated by AI based on user behavior and context. English tech media and professionals have been discussing the trend of "AI agents + personalized distribution," believing that the entry point for information acquisition is shifting from within applications to system-level interfaces.
Meanwhile, Apple and Google are continuously enhancing widget and AI assistant capabilities in their operating system updates and developer ecosystems, attempting to lock user attention at the system level rather than within a single application. This design of "home screen as entry point" is being viewed as a key position in the next generation of traffic distribution competition.
Source: Public Information
ABAB AI Insight
This type of product is not about "better information streams" but rather a complete rewrite of the information production and distribution chain. Traditional social media information streams rely on a three-layer structure: creators produce content, platforms distribute it, and users consume it. The AI agent system directly bypasses the creator layer, turning "content" into on-demand generated function outputs, shifting supply from scarce to infinite.
When content supply is infinite, the focus of competition shifts from "who produces content" to "who defines what the user needs at this moment." This means that information streams transition from "content-driven" to "context-driven," significantly increasing the importance of data loops—device data, behavioral trajectories, time, and scenarios become the new core production materials.
Moving the information stream to the operating system's home screen is a further contest for distribution power. In the mobile internet era, super apps controlled user time through internal feeds; in the AI phase, the entry point is migrating to a lower level, and whoever occupies the home screen is closer to "default choice power." This is consistent with the historical competition for browser homepages, search boxes, and app store recommendations.
A deeper change lies in the attention pricing mechanism. The traditional information stream business model relies on advertising and content distribution efficiency, while AI-generated information streams can match individual needs at a finer granularity, theoretically improving conversion efficiency. This will drive the re-pricing of multiple industries such as advertising, recommendations, and e-commerce, while also strengthening platforms' control over user cognitive paths, further intensifying the concentration of information power.