Even though Google’s NotebookLM was first released as an experimental software application in 2023, it gained wide popularity in September this year after the introduction of an audio generation tool. Google launched on Thursday 17 October, the updated audio generation tool, Audio Overviews, and many more features.
NotebookLM is an AI-powered tool used for researching and document summarization. It has the features that every other generative tool has, but the key attraction is the audio generation. NotebookLM does not just read out the generated texts but sounds more like real human podcasts.
The uploaded files or study materials can be turned into questions and answers to debates and podcasts, which cannot be found elsewhere. Anything can be turned into a conversation, and it feels like two people are there talking to each other. The two AI hosts can read out the summarized materials, and create an interactive session between them. The generated audio can be used to get inspiration and new ideas on a topic.
Audio Overviews are a customizable audio generative tool in NotebookLM. This is sometimes termed as ‘deep dive’. The updated feature allows the podcasts to be customizable giving suggestions, questions, where to focus, and adjusting the level of expertise. The conversion between the AI hosts can now include the users’ presence too as extra suggestions and notes are included other than the uploaded material.
Another update to Audio Overview is enabling background listening. Now the generated audio can be listened to even while multitasking, working on other pages, or continuing the work in NotebookLM itself.
NotebookLM supports Google Slides, links to YouTube videos, and web URLs other than texts and files to create study guides, summarizing notes, and many more. It takes a few minutes more to generate the audio compared to other AI text-generative tools, but have more features to offer. The audio generated can be adjusted to increase or decrease the speed. Currently, the hosts available are one male and the other female, and the language option is limited to English.
The new update can summarize the only topic that needs to be discussed from the large material we have submitted, like the tone and artistic devices from a lengthy novel.
The AI tool uses Gemini 1.5 to generate texts. Even though NotebookLM had a tag, ‘experimental’, Google says that they are collecting user experiences and feedback only, and not any personal data or information. Yesterday, 17 October, Google removed the term experimental after proper research and releasing new and improved features.
This can be taken as a sign of the potential upcoming user retention, better interface, higher software stability, and improved overall quality.
Privacy was given great priority in NotebookLM and it still now promises to keep all the information shared with it secure and not shared for research purposes or any other activities.
Getting access to the generative tool only takes the usual steps of logging in with the Google ID, creating a new notebook, adding the source, a minimum of one, and clicking on Generate. The Audio Overview can be customized with instructions by clicking on Customize. The audio generated is based on the information provided and is not an objective view of the topic.
Google has also announced NotebookLM Business, an upcoming version of NotebookLM with features exclusively suited for a workspace and accessed through Google Workspace. This version would suit businesses, universities, and organizations. Google has also started accepting applications for the NotebookLM Business pilot program to get early access to product features, training, and email support. The pricing and availability will be announced later this year.