Recently, Microsoft began rolling out its new Copilot Notebooks in its Frontier early adopter program. At Synozur, we’ve been exploring Copilot Notebooks over the past few weeks, based on our experience with Google’s NotebookLM.
If you’re exploring AI-powered notebooks, you’ve likely heard of Google’s NotebookLM and Microsoft’s new Copilot Notebooks. Both aim to help you organize and understand information with AI, but they’re at very different stages of maturity. Here’s a casual rundown comparing the two, and why Microsoft’s newcomer isn’t quite on Google’s level yet (though it might catch up soon).
🔎 Google’s NotebookLM – Feature-Rich and Established
Google’s NotebookLM (from Google Labs) has been around a bit longer and it shows in its feature set:
Audio “Podcast” Summaries: NotebookLM can generate an interactive audio overview of your documents. Two AI hosts have a conversation summarizing your uploaded sources (like PDFs or Google Docs) in a lively “deep dive” style. You can listen as if it’s a podcast and even download the audio for on-the-go learning. (I use this sometimes when I know I’ll have “car time” to listen. I notice things differently when I’m listening as opposed to just reading.)
Mind Mapping & Visual Summaries: It also offers Mind Maps that turn key concepts from your sources into an interactive visual diagram. This helps you see the big picture – for example, a student could get a mind map of main topics from their research paper (e.g., “Ocean Acidification,” “Rising Sea Temperatures,” etc. for a marine biology thesis). You can even download the mind map as an image for reference.
Notebook Sharing & Collaboration: NotebookLM isn’t a solo experience – you can share your notebooks with others. It supports adding collaborators via email with view or edit permissions. Real-time collaboration means teammates can work on the same AI-assisted notebook together, which is great for group projects or joint research. (Think Google Docs style teamwork, but in an AI notebook!)
Annotations and Notes: As a true “notebook,” NotebookLM lets you add notes and annotations alongside AI answers. You can jot down insights, highlight text, and save Q&A results as notes. In fact, saved AI responses (like a Q&A) show citations and can be treated like any other note in your notebook. This makes it easy to build a knowledge base with your comments plus the AI’s context.
Broad Content Support: Google’s tool is pretty flexible with what you can feed it. You can upload PDFs, websites, YouTube links, Google Docs/Slides, and more. It then grounds the AI in your content, providing answers with quotes and sources from your materials. NotebookLM even expanded to support Google Slides and web URLs and can create things like study guides or briefing docs automatically.
In short, NotebookLM is ahead in offering multiple ways to consume and organize information (audio, visuals, text), plus letting you collaborate. Your notebooks remain secure inside a Google Workspace. And you can explore free notebooks (with some limitations) along with Google’s paid services.
But it’s got a significant hurdle – especially for the millions for Microsoft 365-centric organizations across the world. You need to convert Office docs to PDF and upload them to Google. This is OK for static information, but for team content that’s being continually updated, your notebook may be out of date without constant manual exports and imports.
🆕 Microsoft’s Copilot Notebooks – Early Days & Potential
Microsoft’s Copilot Notebooks just launched (as part of Microsoft 365 Copilot) in a very early preview state – so it’s much more basic right now:
Core Idea: A Copilot Notebook is an AI-powered workspace where you gather references (documents, chats, meeting notes, links, etc.) related to a task, and then ask Copilot (the AI) questions with that context. Think of it like making a project folder and having an AI assistant read everything in it to give you focused answers. This is awesome for getting tailored, context-aware answers grounded in your own content.
M365 Integration (Strength): One big advantage is its deep tie-in with Microsoft 365 (M365). You can pull content from your emails, Teams chats, OneDrive, SharePoint files, meeting transcripts, etc., directly into a notebook as “references.” Copilot will then use exactly that content to answer you, which is great for enterprise users with tons of internal docs. It’s leveraging your M365 ecosystem — something Google’s tool would do with Google Workspace content. (Using your trusted internal data is a smart way to get relevant AI answers.) This also means your content stays inside the Microsoft 365 trust/compliance boundary.
Lack of Advanced Features (Weakness): Right now, Copilot Notebooks lack many of the cool features NotebookLM offers. For example, there’s no mind map visualization, no timeline, no FAQ or study guides, and no built-in concept of annotating with AI-generated notes. It’s a pretty straightforward Q&A and drafting tool at the moment. Microsoft’s release is very preliminary – effectively version 1.0 with basic functionality. Even simple things like supporting all content types aren’t there: you can only add certain file types (Word, PowerPoint, Excel, PDFs, OneNote/Loop pages) as references, no images, or videos, and not most websites. So, it feels a bit limited for now.
No Team Collaboration (Yet): Despite Microsoft 365’s collaborative nature, Copilot Notebooks are currently a single-user experience. You cannot share a Copilot Notebook with colleagues or co-edit it together – unlike Google’s offering. There’s also no concept of adding an entire team or a shared library of documents in one go; you have to individually add files or references. This lack of sharing is notable, and Microsoft will need to address it to catch up (since brainstorming with a team is a common scenario for using such notebooks).
Built on Loop Components: On a geekier note, Copilot Notebooks are built on Microsoft’s new Loop component framework (stored in SharePoint Embedded, so they stay within the Microsoft trust boundary. Loop is all about real-time, multi-user content, so this under-the-hood tech signals that Microsoft plans to make Notebooks collaborative in the future. Right now, you wouldn’t notice this as an end-user (since, ironically, you can’t multi-edit yet), but it’s a foundation for potential multi-user notebooks down the line.
Rapid Evolution Expected: The good news for Microsoft fans is that Copilot Notebooks is evolving quickly. It’s brand new and Microsoft is rolling out updates constantly. In fact, some features inspired by NotebookLM are already being tested – for example, Microsoft recently announced an “AI podcast” feature and deeper research tools for Copilot (sounds a lot like NotebookLM’s Audio Overviews and advanced querying)[2][2]. They also introduced Copilot Pages (a canvas to organize research, similar to a scratchpad)[2]. These are just starting to appear for some users in preview. So, Microsoft is basically saying “we see those gaps and we’re on it.” 🚀
⏩ What to Expect in the Next 6–12 Months
Microsoft is very likely to close the feature gap within the coming year. Given the resources and focus behind Copilot, expect rapid improvements: collaborative notebooks, support for more content types (images, web pages), and richer outputs like visual summaries or audio, should all arrive in time. Microsoft has already hinted that many of NotebookLM’s best ideas are making their way into Copilot. So, while today Google’s NotebookLM is clearly ahead in functionality, Microsoft’s Copilot Notebooks is a fast learner that could catch up sooner than we think. Competition in this space is heating up, which is great news for us users.
🎉Bottom line: Copilot Notebooks (Microsoft) is a promising new kid on the block but still playing catch-up to NotebookLM (Google) in late 2024/early 2025. If you need mature features right now (like interactive audio summaries, mind maps, or easy sharing), Google has the edge. If you’re invested in the Microsoft ecosystem, keep an eye on Copilot Notebooks as it rapidly evolves – it might surprise us with how much it can do in the next year.
Hope this comparison helps! Feel free to share your thoughts or experiences with either tool. 🤖📓✨