NotebookLM becomes Gemini Notebooks
Hans Lucas/AFP via Getty Images
At Google I/O in May, I lost track of whether Google was announcing new products or renaming old ones.
The announcement of Google Pics was a good example of this. Was it an evolution of Google Photos? A replacement for Google Images? Neither. It was a new tool that sounded similar and also overlapped on some functionality. I was confused. To be honest it doesn’t take much.
Today Google has renamed NotebookLM to Gemini Notebook. At the same time, Gemini already has notebooks. A notebook inside Gemini is not quite the same experience as the standalone NotebookLM product, now Gemini Notebook, but the two are connected and synchronized. So are they different products or do we now have Gemini Notebooks and Notebooks in Gemini. I’m confused again.
But I think renaming NotebookLM is ultimately a good move.
The Tool Formerly Known As NotebookLM
I think NotebookLM, sorry Gemini Notebooks, is the most useful AI tool available to educators.
I start almost every workshop with it. Teachers and leaders tend to understand its value quickly. They bring their own material, ask questions about it and follow citations back to the original sources. It feels less like asking a chatbot to know everything and more like working through documents with a capable research assistant.
I have seen educators use it to build courses, examine school data, learn unfamiliar topics and turn existing material into resources for students. The uses vary because the starting point is whatever the educator already has.
That helps explain why Google says more than 30 million people and 600,000 organizations now use it. Audio Overviews widened the audience by turning source collections into conversations between two synthetic hosts. Strange at first, but often useful.
Notebooks With Cloud Computer Access
Google is also giving each notebook access to a secure cloud computer. Gemini Notebook can write code, run it and use the results to analyze the sources inside a notebook. It moves the product from helping people read research to carrying out parts of that research.
A school leader could upload attendance, assessment and behavior records, then ask the notebook to clean the data and look for patterns. A teacher could compare curriculum documents across several subjects. Someone evaluating a new program could combine survey files, produce charts and test whether a pattern holds across groups.
Access begins with Google AI Ultra users and eligible Workspace business customers. Google says Pro users will receive the capability on the web over the coming weeks.
There is a catch. The easier analysis becomes, the easier it is to accept a result without understanding how it was reached.
Researchers studying code produced by six language models found that many errors stretched across several lines and required substantial repair. Some incorrect programs still ran. For a school leader who never sees the code, a convincing chart could possibly hide a poor decision about missing data, categories or methodology.
The Education Question
This arrives when student AI use is already routine. The Higher Education Policy Institute surveyed 1,054 full time UK undergraduates in late 2025. It found that 95% had used AI in some form and 94% had used generative AI for assessed work.
Yet fewer than half believed teaching staff were helping them develop the AI skills they would need in their careers. HEPI policy manager Charlotte Armstrong said “AI literacy and capability must be embedded across the curriculum.”
Gemini Notebook could complicate that task. If a student asks it to tidy a dataset, perhaps that creates more time for interpretation. If the system selects the method, writes the code, runs the analysis and explains the result, what exactly has the student learned?
That question comes up repeatedly in my workshops. Educators see the time saving potential almost immediately. Then someone pauses and asks which part of the thinking should still belong to the learner.
Schools need to be precise. Students should learn to question the method, inspect assumptions and check important findings against the original material. The skill is no longer simply doing every step unaided. It is knowing which steps can be delegated without surrendering judgment.
Google may have confused me with it’s naming practices, but it has made the product considerably more useful.
NotebookLM helped educators talk to their sources. Gemini Notebook can now start working on them.