NotebookLM : Article to Podcast as a Magic
NotebookLMs audio overview feature is quite amazing on how it can convert any research article, even a boring one into an interesting one.
I recently came across something about NotebookLM that completely caught me off guard. Apparently, this tool isn’t just for taking notes anymore—it now transforms any article into a podcast. 🎧
At first, I shrugged it off, thinking it was just another text-to-speech tool like ElevenLabs or Deepgram or Whispr by Open AI. I was wrong. When I gave it a try, I was genuinely blown away. It’s not just “text-to-speech”; it’s a whole new level.
For those who haven’t heard of NotebookLM, let me break it down. It’s a product from Google Labs, essentially your personal AI research assistant, powered by Google's Gemini 1.5 Pro model. Think of it as having a virtual research buddy. You upload documents or notes, and this thing instantly becomes an expert on your content.
But here’s what got me really excited: Audio Overview. This new feature doesn’t just read your article back to you—it converts it into a full-blown conversation between two voices. It’s so engaging that even the most boring research papers suddenly feel like an interactive story. 🙌
Here’s how you can try it:
Go to NotebookLM.google.
Upload that article or document you’ve been procrastinating on.
Click on Audio Overview and watch the magic happen. 🪄
I even asked NotebookLM what kind of use cases it’s best for, and its AI spit out some impressive answers.
Here are some of the best use cases for NotebookLM written by NotebookLM itself:
NotebookLM for Research: NotebookLM is helpful for research as it can synthesise information from multiple sources, such as PDFs, Slides, Google Docs, text files or web pages. NotebookLM can act as a virtual research assistant, able to answer questions from uploaded source material.
NotebookLM as a Help Center or Onboarding Guide: NotebookLM can be used to create help documentation, acting as a guide for new users of a product or organisation. NotebookLM is able to answer both factual questions from provided sources and more freeform queries. For instance, you might ask NotebookLM "How many words can I include in a source?" or "I’m trying to put together a legal brief that draws on several existing court decisions. How can NotebookLM help with that?".
NotebookLM with Meeting/Interview Transcripts: NotebookLM allows you to upload meeting and interview transcripts and ask questions based on their content. For example, you could upload earnings call transcripts and ask, "What did Bob Packman say about profitability in Q3 of 2022?". You could also upload interview transcripts and ask, “What are some controversial quotes about the Homestead Strike of 1892” or “Who was it who mentioned the metaverse and what was their point about it”.
To get the most out of NotebookLM, ensure each paragraph of the transcript names the speaker, especially for transcripts longer than 20,000 words. This helps NotebookLM to track who is speaking. For example:
Bob Smith: It’s great to have everyone here for this meeting. Thanks for coming. Let’s begin by discussing the marketing plan for 2024. It’s going to be a very big year for us, and there’s much to discuss. Bob Smith: I think the biggest issue we face is our advertising spend on television. Julie, can you discuss this in more detail.
Julie Abrams: Thanks, Bob. Happy to dive into the numbers.
You can also add time codes at the beginning or end of paragraphs, which helps if you need to ask a question about a specific time, for example:
Bob Smith: It’s great to have everyone here for this meeting. Thanks for coming. Let’s begin by discussing the marketing plan for 2024. It’s going to be a very big year for us, and there’s much to discuss. (1:03:15)
NotebookLM for Reading: NotebookLM can be used to improve reading comprehension and retention. When you upload a source, NotebookLM automatically generates a source guide with a summary and a list of key topics. You can also ask NotebookLM suggested questions to further explore the source. NotebookLM allows you to highlight passages of text and provides suggested actions, such as saving interesting quotes as notes. You can also ask NotebookLM to summarise passages or translate them into simpler language.
Notebook LM for taking Notes: NotebookLM offers two types of notes:
Saved Responses: You can save useful responses to questions or instructions as notes. Saved responses often include citations, allowing you to easily locate the information in the original source.
- It's important to note that citations in saved responses are formatted slightly differently and are limited to 10 per note.
Written Notes: You can write your own notes in NotebookLM. You can then use your written notes as a source for NotebookLM, allowing you to ask questions based on their content.
NotebookLM for "Curate and Create": You can curate a collection of notes by pinning important information to the noteboard. These notes can be saved responses, quotes from your sources, or your own written notes. Once you have a collection of notes, you can select them and ask NotebookLM to perform a variety of actions:
Combine notes into a single note
Summarise the contents of the notes
Organise notes into thematic categories
Create a study guide from the notes
Create a custom format, such as a blog post, marketing plan, or email newsletter.
NotebookLM for Writing: NotebookLM has a range of tools to help with your writing style and substance. You can write notes directly in the app, or use the 'curate and create' function to generate an outline or a draft of an opening paragraph for your writing. NotebookLM also allows you to select a written note and ask for a critique of your argument or for suggested related ideas based on your writing.
NotebookLM still has its limitations as we cannot change the characters, the voices, not modify the content in this. I am sure, that will change in days to come. However, for now, it’s quite amazing and a must-try.
I have made one recently for my article on cursor and it’s fun to hear.
Check it out here.