According to Digital Trends, Google has launched a new feature for its AI-powered notebook app, NotebookLM, called Data Tables. This tool automatically pulls information from a user’s scattered notes, PDFs, and documents and organizes it into structured tables. These tables can then be exported directly to Google Sheets or Docs for further editing. The feature is available now for subscribers to the Google AI Pro and Ultra tiers, with a broader rollout to all users planned over the coming weeks. It’s designed to handle tedious manual organization tasks, like turning meeting notes into action item lists or building competitor comparison charts.
The real workflow shift
Here’s the thing: the magic here isn’t just the table generation. It’s the plain language command. Instead of you figuring out the schema—what should be a column, what should be a row—you just tell the AI what you want. “Make a table from these three research papers comparing their methodologies, sample sizes, and conclusions.” That’s a huge cognitive load off. You’re describing the outcome, not doing the manual labor of sorting and pasting. But there‘s a catch, and it’s a notable one. The tables inside NotebookLM itself aren’t interactive. You can’t tweak a cell. You either export to Sheets to edit, or you go back, tweak your prompt, and generate a new one. That feels like a bit of a half-step, honestly. Why not let me adjust it right there?
More than just notes
This move is a clear signal of Google‘s ambition for NotebookLM. It’s not trying to be just another note-taking app. With Deep Research mode and now Data Tables, they’re building a full research and synthesis workspace. They want it to be the place where messy information goes in and structured, actionable intelligence comes out. For students, researchers, analysts, or even project managers drowning in meeting transcripts, this is a compelling proposition. It turns a passive repository of info into an active analysis tool. And by baking the export directly into Sheets, they’re playing to their ecosystem strength. They’re not just making a table; they’re making a *Google* table, ready for collaboration.
The AI assistant evolution
Look, we’ve seen AI that can summarize and answer questions. The next frontier is AI that can structure. That’s what this is about. It’s moving from a chatbot that responds, to an analyst that organizes. The real-world examples Google gives—action items from meetings, competitor grids, aggregated research—are all classic knowledge worker pains. This is Google saying AI can do the “grunt work” of research. But it also raises a question: how good is it really? If your source documents are a chaotic mess, will the AI’s table be logically sound? It’ll be interesting to see user feedback, like from early testers such as Manisha, on its accuracy. If it works well, it could save a ton of time. If it’s flaky, it’ll just create more work fixing its mistakes. Basically, its utility lives or dies on reliability.
