Notebook LM

Google Notebook LM

Google Notebook LM is, without a doubt, one of the best artificial intelligence applications we know. If you don't know it, you have to get to know it because it can help you in your daily life in many things. We have talked about it at length on this channel occasionally, but we have never done what we are going to do today, which is a video exclusively dedicated to understanding this tool, why it is so important, and how it can help you, in addition to making a step-by-step guide on how to use it and get the most out of it. Stay because this will undoubtedly interest you, whether you have been using it for a while or if you don't know the tool that I assure you will blow your mind.

Introduction and Concept

Notebook LM, as you can see, is an application that comes from Google, and the first thing you can do is use it for free. If you connect to notebooklm.google.com, you will find something similar to this where you have what we call notebooks. Notebooks are those panels, those boards where we are going to share information with Notebook LM.

Origin and Philosophy of Notebook LM

But first, let's go in a bit to understand what this is about and why Notebook LM exists. Because you know that Google's artificial intelligence is Gemini. Well, Notebook LM was born as an absolutely side project made by five guys at Google with no intention of this actually becoming what it has become, but it worked out well for them.

The maximum premise was to create an artificial intelligence that, instead of being anchored in its training or the internet, was anchored in documentation—something we gave it—and from there, it would only work on those documents. Something that, at the time Notebook was born, was very complicated to do with a general LLM because they were all very contaminated by training or information searches, causing them to pull information that mixed all those sources when you gave them a PDF, rather than simply limiting themselves to the information in that document you had uploaded. Notebook LM today goes much further than that, but that remains the principle and the differentiating characteristic. You have an artificial intelligence that will only rely on the documentation and information that you want it to have.

Use Cases and Workflow

But Notebook LM goes much further than that. Let's first look at these general notebooks that exist, which are public, so you can understand a bit of the flow of what this is about.

Source Grounding: The Case of The Economist

Look, if we go inside Notebook LM, we go to the Featured Notebooks section, which are basically the ones they recommend. And I'm going to look, for example, at this one from The Economist, which is an important American newspaper. Inside it, we will see that they have introduced a lot of information sources, which is what you see on the left. The sources are always in the left panel. As you can see, it has many.

And from here it tells us it talks about how the world will be from 2026 onwards. This allows us to have a chatbot trained to talk about 2026, but it will only refer to these documents that have been shared with it, which are documents created by the newspaper itself, and therefore they control the flow of where the information comes from. This allows me, for example, to come here and ask it how Trump's policies will affect the economy and trade, and from there it answers me, but instead of doing it however it wants or based on internet information, it will be based 100%, as I say, on that truthful and important information that the creator of the notebook themselves introduced.

If you notice, it starts to answer, "Donald Trump's policies represent the most radical economic experiment in the United States, ta ta ta," and it lists all the sources, which document and article that part belongs to, and where it got it from. And from there we can always see what it is supporting itself on. All these numbers 1, 2, 3, 4, 5, etc., and all these points always refer to those 70 documents that Notebook LM had introduced. This in itself is already tremendously useful, but as I say, this is what it was initially conceived for.

Capabilities as a Knowledge Converter

Notebook LM has grown a lot, evolved a lot, and I think it has even changed the purpose of the tool. What is Notebook LM for today? And why is it relevant that you know it? And why do you have to know how to use it? Because Notebook LM is a knowledge converter. Basically, we have to understand Notebook LM as a platform where we input information in one way and we can output it in a very different way, very easily and legitimately, in the sense that it will always be supported by those sources.

This allows you to convert things, for example:

• A list of reports you have on a client into a proposal for the client.
• If you have a lot of customer service information from your chats, convert it into statistics on how satisfied your customers are.

It is an information converter with artificial intelligence totally supported by the documentation you upload. I repeat, that is the most important thing, but in the end, you can use it for countless use cases, from studying with it to using it for work, to applying it in your personal daily life.

Getting Started Guide and Usage Plans

First, as I was saying, go to

notebooklm.google

and you will see how you immediately have this page where you can register and link your account. This is totally free, although it has a payment plan for those who want to go further, but I tell you it is perhaps one of the applications where there is the least need for a payment plan. Truly, the free version is very comfortable and extensive in its use, so you don't need to pay.

Limits and Subscriptions

From here you can see the different plans there are, that you have Notebook LM within your Gemini Plus plan because it is included in the account you already have for Gemini. You also have it in the Pro plan and the Ultra plan, and from there the amount of use you can make and the amount of sources you can give it changes completely:

• Free accounts: Up to 50 sources.
• First-level paid plans: 100 sources.
• Second-level plans: 300 sources.
• Ultra plan: You can put up to 600 sources for each of the notebooks.

Setting Up a New Notebook

When you enter, you will see that you can create a new notebook or access one you have already created. Let's create a new one, and the first thing it will suggest is that we give it information. Of course, since this is based solely on the information it has, if it doesn't have any information, the tool is absolutely useless.

Information Sources and Upload Methods

We have many ways to fill it:

• Fast Research or Deep Research: It searches using a fast or deep research like the ones in ChatGPT or Gemini. That is, it launches itself onto the internet to look for information on a topic and compiles it itself. You can even indicate where you want it to look, whether on the web or in your company's Drive.
• Your Own Files: The real potential is in giving the information ourselves. It admits:

PDFs, images, documents, audios. CSVs, Excel documents. Videos, MP3s. Web pages or YouTube URLs (videos require direct transcription). Pure text copied directly.

Data Processing

{IMAGE_PLACEMENT prompt="A high-tech digital interface showing multiple document icons like PDF, CSV, and MP3 being uploaded into a central glowing Google NotebookLM core, with progress bars and a professional aesthetic."}

When uploading files, the system processes them, reads them, embeds them, and summarizes them to understand them. This usually takes a few minutes depending on the number and size of the documents.

Interaction and Control Panel

Once processed, the system shows a summary of the general topic. For example, in a case of geopolitics and AI, the system identifies White House reports, AI action plans, economic leadership, and international tensions. This allows working with valuable information provided by us.

The interface is divided into:

• Left panel: Source management.
• Central panel: Chatbot to interact with the documentation.
• Right panel: The "Studio" block to create new content formats.

Advertisement Section: Cyberghost VPN

Notebook LM is incredible, but in some Google applications like Pomeli, you might find regional restrictions. For this, it is recommended to activate Cyberghost VPN with a location in the United States. Cyberghost encrypts your data, keeps you anonymous, and allows access to extensive streaming catalogs. It has more than 11,000 servers in 100 countries and allows connecting up to 7 devices with a single subscription.

Study Block: Format Conversion

The study block is what is really important because it allows us to create other ways of consuming that content. It allows converting textual information into optimal formats for daily work.

Custom Data Tables

It allows extracting data from the documentation and organizing it into explanatory tables.

• You can choose the language (Spanish, English, etc.) regardless of the source language.
• You can customize via prompts what data is relevant.

This saves hours of manual work in unifying multiple sources.

Infographics and Visual Generation

Using Nano Banana Pro, Notebook LM can generate complete infographics.

• Formats: Horizontal, vertical, or square.
• Detail level: Concise, normal, or detailed.
• Customization: You can request a specific style (e.g., blue and gold colors).

{IMAGE_PLACEMENT prompt="An infographic in professional blue and gold tones showing global maps and AI technology icons, structured vertically with clean typography and technical diagrams about geopolitics."}

Unlike a conventional GPT, Notebook LM strictly respects the original source and does not contaminate the result with external information from the model's training.

Study Tools: Quizzes and Flashcards

• Quizzes: Generates tests based on the uploaded syllabus.
• Flashcards: Interactive cards with animated questions and answers for memorizing content.

Reports and Automatic Summaries

It allows generating:

• Briefings.
• Study guides.
• Blog posts.
• Case studies.

The system automatically suggests the most useful type of summary based on the detected documentation.

Advanced Features: Multimedia and Presentations

PowerPoint Presentations (Video Overviews)

It allows creating narrated presentations or PowerPoint files.

• Slide types: "Presenter slides" (clean for exposition) or detailed (to send by email).
• Length: You can define how extensive the presentation should be.
• Style: Visual design can be defined using natural language.

Each slide is generated as a full image by Nano Banana, allowing for total visual coherence.

Slide Editing and Review

The "review" button allows modifying individual slides using natural language:

• Change visual elements (e.g., change a map).
• Remove logos or specific elements.
• Add additional text points.

The system regenerates the slide image to meet the new requirements. Currently, they can be downloaded as PDF or PowerPoint (as static images).

Audio Overviews: The Podcast Format

Generates a conversation between two people (AI voices) about the provided information.

• Formats: Deep dive, briefing, critique, or debate.
• Utility: Allows consuming complex information in situations where you cannot read (driving, gym, etc.).
• Interaction: In the English version, it allows interrupting and interacting with the speakers to go deeper into specific topics.

Regional Settings and Interaction with Gemini

It is essential to set the "Output language" in the settings to ensure that the generated content is in the desired language.

Notebook LM allows sharing notebooks at a business level, defining if the receiver has access to all outputs or just the chat. Additionally, there is an integration with Gemini: when adding files in a Gemini conversation, an existing notebook can be selected so the AI has access to that specific context as a reference source.

Conclusion and Additional Resources

Notebook LM is one of the most powerful tools in the Google ecosystem, even if it is not yet fully integrated into Gemini. It facilitates productivity by allowing the management of up to 600 sources of truthful and personal information.

For more information, a guide with 20 use cases and step-by-step tutorials is available in the links of the original video description.