Google is currently testing two cutting-edge features, "Personal Intelligence" and "AI Note Editing," in its research and note-taking tool NotebookLM, indicating that the product is accelerating its evolution from a single document reader to an adaptive personal intelligence collaboration center.
Previously, Google had made a major architectural upgrade to NotebookLM in early June 2026, switching its underlying dialogue model to Gemini 3.5 Flash, introducing Antigravity-driven software technology, and providing a dedicated cloud code execution environment for each notebook, significantly enhancing the ability to handle complex tasks and improve output quality.
The newly exposed test features closely align with two industry trends: "deep collaboration" and "long-term memory." The "Personal Intelligence" feature gives NotebookLM the ability to retain context across conversations, allowing the system to store and refer back to past interactions. To meet the privacy needs of research and enterprise users, this feature allows users to manually turn it off and review stored content, with data interactions strictly limited within the NotebookLM application and not integrated with other Google services such as Gmail or Docs.
The other feature, "AI Note Editing," closes the loop between the canvas and the conversation, allowing users to select note text directly and push it into the chat window as context for AI to rewrite and refine, completely solving the pain point of being unable to modify results after they are generated.
The exposure of these two features not only represents a proactive response to the mainstream AI tool trend of "memory and personalization," but also marks Google's deeper integration of NotebookLM into its large model ecosystem. As competition in AI research tools intensifies, by reducing the collaborative friction between drafting and questioning, NotebookLM is gradually building higher user engagement and professional barriers.
