by: Jeff Muday, IT Analyst, Biology and Chemistry

Summary

NotebookLM is a “free”, AI-powered research assistant from Google that helps users organize, analyze, and create outputs from their research materials. It utilizes Gemini, a large language model with a 1.5 million token window. NotebookLM allows users to synthesize and interact with large amounts of content and can create notebooks using various input methods, like Google Docs, links, and pasted text. It supports up to 50 sources in formats such as PDF, text, markdown, and audio, and can process audio from YouTube videos.

At the risk of repeating myself, I mentioned earlier that one of the major features of NotebookLM is that it has a HUGE context window far exceeding the current ChatGPT, Claude, or Deepseek models. This means you can load up to the equivalent of around 1.5 million words, enabling it to provide detailed and accurate answers based on a vast pool of source material.

We have witnessed many past Google projects discontinue or fail– for example, their Google Wave, Google Plus, etc.  I am a bit wary of the long-term nature of NotebookLM, but I think this one may have some traction.  They have a nascent pricing model for their Plus version, but you can get started for free.

How to Get Started:

1. Sign in to Google to access NotebookLM with a Google account.

The URL is: https://notebooklm.google

2. Create a new notebook.

3. Upload documents like PDFs or Google Docs into the platform. The Sources tab will support up to 50 documents. Other options for sources would be YouTube videos (it automatically transcribes the audio) or Website URLs. (Note that it only loads the current page/article of a website, you may have to include more URLs to capture your topic fully.)

4. Now that you have a source, you will see an executive summary:

5. You can now ask questions or generate summaries by interacting with the AI.

For example, I can ask the question, “What Data Science contributions were made to this project?”

6. You can create notes by pinning important information and creating notebooks.
7. “Audio Deep Dive” or “Audio Overview” creates a nice podcast with fluent natural voices! Students may find this a compelling option if they are auditory learners. Simply click “Audio Overview” and it will create a podcast summary.

8. Check out the “Studio” tab for more information. While you are waiting for your podcast to be generated, you can create a “FAQ”, “Study Guide”, “Timeline”, etc.  Lots to look into here!

Pros and Cons of using NotebookLM

Pros of using NotebookLM:

1. Fast and accurate summaries of lengthy documents. NotebookLM offers AI-driven summaries, insights, and the ability to answer complex questions based on uploaded documents.
2. Custom queries help pinpoint specific information, and contextual search allows users to query documents with custom searches. Users can also create custom queries to tailor the AI’s responses to specific information needs.
3. Source citations ensure accuracy and reliability. The citation feature links information to specific sources.

Cons of using NotebookLM:

1. Limited formatting options for uploaded PDFs.
2. Pricing is still in flux, even though they have a Free version and pricing for the “Plus” tier, it is still in the experimental phase.
3. Occasional inaccuracies in document summaries require fact-checking. It may sometimes flatten or exaggerate differences between documents and is still prone to some hallucinations and misinformation.

Should I Be Using NotebookLM

Notebook LM simplifies research and note-taking by summarizing documents and generating insights. It is a valuable tool for managing extensive documents, extracting key information, and organizing complex content. However, it is essential to remember that the tool is still experimental and may have limitations.

Although NotebookLM is private, remember to avoid sensitive information or typing sensitive queries that you would not want others to see.