01What Are the NotebookLM Updates?

NotebookLM updates transform the tool from a simple question-answerer into an AI that does real work. Instead of just reading your files and replying with text, it now has a secure cloud computer that can write and run code. That means it can crunch numbers, clean data, and build finished documents for you.

In this episode, Julian Goldie explains how Google gave NotebookLM its biggest update ever. The headline is that every notebook now comes with its own secure cloud computer. This computer runs in the background and can perform tasks, not just talk about them.

So if you give it a messy spreadsheet, it can clean it and build a chart, all on its own.

02Main Features in the New NotebookLM Update

This update is packed with new features. Julian highlights several key changes that make NotebookLM far more powerful than before.

  • Secure cloud computer: Each notebook gets a real computer that can write and run code, so it can do tasks like math and data cleaning.
  • Gemini 3.5: The tool now runs on Google's newest and smartest model, making answers more accurate.
  • Antigravity system: This system lets you see how NotebookLM thinks, showing its reasoning step-by-step, so you don't have to trust a black box.
  • Over 100 software skills: Built-in skills include cleaning data, pulling key facts, finding research gaps, and spotting connections.
  • Web access: You can start with just a rough idea, and NotebookLM will search the web to find real live information for you.
  • 11 output formats: You can download finished files, including PDF, Word, PowerPoint, Excel, CSV, JSON, charts, and images.

03How to Get Started with the New NotebookLM

Getting started with the new NotebookLM is easier than ever, because you don't need to have everything prepared. Julian explains that you can walk in with just a loose question.

  1. Open NotebookLM and create a new notebook.
  2. Type a rough idea or a question about the topic you want to understand.
  3. Let NotebookLM search the web for sources, or upload your own files like PDFs and spreadsheets.
  4. Mix the web sources with your own materials to give NotebookLM a broad base of information.
  5. Ask NotebookLM to perform a task, like cleaning data or building a report.
  6. Wait for it to use its cloud computer, and then download the finished file in your chosen format.

The key is that you don't need to be organized. You just need to be curious and ask.

04How to Use NotebookLM's New Features in Real Work

Julian shows how the new NotebookLM can take over real tasks. For example, say you run a small business and have three years of messy records in a spreadsheet. You can hand it to NotebookLM and say, "Clean this up and show me what's going on." It will clean the data, find patterns, and build a chart for you.

Another example is growing a community. You can feed NotebookLM your notes, member feedback, and research on what people want. Then ask it to build a one-page report showing the biggest benefits people get from joining, with charts.

It will produce a clean PDF ready to share in minutes.

Because it has web access, you can start with a vague topic and let it gather sources. Then it mixes that with your own files. You can also ask for different output formats, like a PowerPoint presentation or an Excel budget sheet, and it will build them for you.

05Results and Comparisons from the Episode

Julian doesn't run formal tests, but he shows the difference between the old and new NotebookLM. Before, it could read a spreadsheet and tell you what was in it, but it couldn't crunch the numbers. Now, it opens the cloud computer, writes code, runs the math, and hands you the answer.

He also compares the output. Old NotebookLM gave you text in the chat, like summaries. You had to take that text and build a report or slide deck yourself.

Now it builds the finished thing in 11 formats. For instance, you can ask for a PDF report with charts and tables already inside, or a PowerPoint with real slides.

Julian emphasizes that these are not pre-made templates. When you ask for a file, NotebookLM plans it, writes the code, runs it, and delivers the final product. This turns tasks that used to take a full afternoon into a single sentence and a short wait.

06Tips, Limits, and Mistakes to Avoid with NotebookLM

Julian is straightforward about the catch. This big update is not free for everyone yet. Right now, it's rolling out to people with a Google AI Ultra subscription and to business customers on certain Workspace plans.

If you're on the free version, you still have the older NotebookLM for now. But he notes that free tools always catch up, so it's smart to understand the new features early.

Another tip is to avoid staring at a blank page. The old way required you to have your files ready, but now you can start with just a rough idea. Use the web search feature to pull in sources.

That way, you don't get stuck.

Also, remember that you can edit any of the files it creates. So even though it builds the finished product, you can tweak it to make it yours. That's useful for client-ready documents.

Finally, don't wait. Julian stresses that the gap is between people who learn tools early and those who wait. By understanding the new NotebookLM now, you'll be ready when it reaches your plan.

07Who NotebookLM Updates Are For and Why They Matter

These updates are for anyone who spends time on grunt work—reading long documents, cleaning data, turning notes into reports, or building slides. Julian points out that this kind of work takes hours every week and none of it is the fun part. NotebookLM takes most of that off your plate, so you can focus on decisions and strategy.

It matters because the AI doesn't just answer anymore. It works. Julian says the thing you used last month and the thing that's live today are not the same product.

This is the pattern with AI tools right now: they jump from doing one job to five jobs in a single update.

People who learn these tools early get their work done in a fraction of the time. They handle more clients and projects with the same hours. If you want to avoid falling behind, understanding NotebookLM's new capabilities is the smart move.

It's a brand new tool wearing an old name.

08The Bigger Picture of NotebookLM's Evolution

Julian places this update in the bigger trend. NotebookLM used to be a reader. It read your stuff and told you about it.

Now it's a doer. It reads, thinks, browses, builds, and delivers a finished file. That's a completely different tool.

He says this is the clearest sign yet of where everything is going. The AI doesn't just answer anymore. It works.

The tools are jumping forward so fast that people who stay updated pull ahead of those still learning the old version.

To put it to work, Julian suggests joining communities like the AI Profit Boardroom or the AI Success Lab. But even on your own, you can start experimenting with the new features now. The key is to understand how this works so you're ready the second it's in your hands.