Google on Thursday announced new updates to its AI note-taking and research assistant, NotebookLM, allowing users to get summaries of YouTube videos and audio files and even create sharable AI-generated audio discussions, as the search giant aims to broaden the use cases and reach of the tool, which was originally launched as a project at its I/O developer conference last year and expanded to markets including India, U.K., and more than 200 countries months after its debut for public access in the U.S.

NotebookLM, which was initially used by educators and learners, has more recently seen a significant shift in its user base and is now attracting more people in a workplace environment.

Raiza Martin, a senior product manager for AI at Google Labs, said in an exclusive interview that the tool’s users are now roughly split, with 50% being educators and learners, and the other half consisting of business professionals.

“People are now sharing notebooks, and it’s creating a network effect,” she said.

This gave the NotebookLM team a push to introduce new features, with the hopes of enhancing its network effect and making the tool popular among people of different demographics.

Earlier this month, NotebookLM added Audio Overview to let users turn their documents into engaging audio discussions. The latest update expands that experience by allowing users to share the Audio Overview generated in NotebookLM with a public URL.

To use the feature, you click on the share icon available on the Audio Overview generated in the tool to get its URL, which you can then copy and share with others.

Martin said that her team saw professionals uploading web pages, resumes, and even their presentations on NotebookLM to generate Audio Overviews, then sharing those with their employers, colleagues, or clients.

NotebookLM also added support for YouTube videos and audio files (such as .mp3 and .wav) as new source types, alongside its existing support for Google Docs, PDFs, text files, Google Slides, and web pages. The new features help users summarize key points from YouTube videos and generate takeaways and insights from the audio recordings they have of their study sessions or projects.

Martin said that because Google Labs has a small team working for NotebookLM, powered by the company’s multimodal large language model Gemini 1.5 Pro, every new feature the team adds to the tool is based on user feedback.

“What’s interesting about AI tools is that a lot of assumptions change,” she said. “What might have been useful last year might not be useful this year.”

In June, Google expanded NotebookLM access to more than 200 countries after launching it initially in the U.S. late last year.

Martin said, without disclosing specific numbers, that while NotebookLM still has a majority of its usage in the U.S., Japan has emerged as the next big market for the tool. The executive also underlined that some of its users are using NotebookLM to get AI-based summarization in languages different from what they set on the tool.

“In Japan particularly, we see a lot of documents that are not in Japanese, but NotebookLM is set to Japanese,” she said. “So people are querying in their native language, using it with probably complex and dense documents in English.”

Google said that the information users upload to NotebookLM stays private and is not used to train an AI model. To access the tool, users must be at least 18 years old.

Still, NotebookLM faces challenges inherent to its nature as an AI tool. One is that if users rely too much on NotebookLM, they might quickly lose the habit of reading long-form content and research papers. This could also lead to the problem of oversimplification.

Martin said that her team is very well aware of these concerns.

NotebookLM provides clickable citations from the content users upload to let users go deeper into the summarized notes.

“We try to encourage you to read your original text. We encourage you to double-check all the answers that come out of NotebookLM… You could read SparkNotes or the actual book; it’s always up to you,” she said.

NotebookLM is currently limited to the web, though Martin indicated that its mobile apps may come sometime next year.

Meanwhile, the team is busy adding more new features. These will be focused on adding more support on the input side and new sources for outputs, Martin said.

Google announced on Wednesday that its AI note-taking and research app, NotebookLM, is adding an “Audio Overview” feature. Audio Overview will give users another way to digest and comprehend the information in the documents they have uploaded to the app, such as course readings or legal briefs.

Since its launch, NotebookLM has used text to summarize and explain source materials, but now, it can do so out loud using audio. The feature is geared toward people who grasp materials better by listening to explanations as opposed to reading them.

AI-generated virtual hosts will use conversational speech patterns to give you a summary of the materials you have shared. The hosts will share facts or compelling topics from the source material and do things like use metaphors to explain complicated concepts. Google says that listening to these discussions will help users find new connections between their documents or get inspiration for their drafts.

To get started with Audio Overview, you need to open an existing notebook, navigate to the Notebook guide, and click on the “generate” button on the right-hand side. It can take up to five minutes for an audio conversation to generate, and for now, you can only generate conversations in English.

You also have the option to download the conversation and listen to it on the go to get a quick understanding of your source materials.

Audio conversations are still in beta and there may be some inaccuracies in the conversations, Google says.

The company first demoed NotebookLM under the name Project Tailwind back in June 2023 at Google I/O before making it available to everyone above the age of 18 in the U.S. last December. The app is now available in over 200 countries in more than 108 languages.