Google’s NotebookLM, got a big upgrade, giving users more control over its AI-generated audio conversations in podcast style. NotebookLM, initially rolled out as a tool for summarizing information, now lets people guide these summaries in ways that better match their needs, particularly when creating audio overviews. The feature upgrade comes alongside a new business pilot aimed at organizations.
New Level of Audio Customization
Before, NotebookLM would automatically create audio from the uploaded documents, which some found limiting. With this latest update, Google introduced a “Customize” button, letting you tell the AI exactly what to talk about. For instance, you can get the conversation to focus on a specific part of your content, cutting out irrelevant bits. According to Google, this makes the tool more useful for researchers and students who need accurate and focused discussions based on their material.
Customizing the audio not only keeps the conversation on track but also helps prevent “hallucinations”—those moments when AI makes up information not actually in the source.
Work While You Listen
NotebookLM also added a feature that allows you to keep working while listening to the AI-generated summaries in the background. Now, instead of stopping to listen to the entire summary, you can query your sources, gather citations, or explore quotes without pausing the audio. This improvement turns NotebookLM into a more flexible multitasking tool, particularly for people juggling large datasets or extensive research.
The tool has grown significantly since its public debut in December. Initially targeted at students and researchers, it’s now catching the attention of businesses as well. Google expanded NotebookLM into over 200 countries, and it’s seeing around 4.17 million monthly visitors, according to traffic data from SimilarWeb.
Introducing NotebookLM for Businesses
In a major shift, Google is also launching a business-focused version of NotebookLM. The NotebookLM Business pilot introduces advanced capabilities tailored to organizations, including higher usage limits and the ability to share notebooks among team members. Participating companies will gain early access to these new features, alongside personalized training and email support.
The NotebookLM Business pilot is currently open to applications, but there are no details yet on its general availability or pricing structure. Google did say that these business-focused features will offer a more collaborative approach, making it easier for teams to manage collective knowledge and streamline project onboarding.
This pilot will help Google gather feedback from organizations, allowing them to tailor future developments for business needs.
Expanding Sources and Global Reach
NotebookLM isn’t just about text files anymore. Google recently added support for YouTube videos and audio files as sources for generating summaries. Previously, the platform worked with PDFs, Google Drive files, and URLs. Now, these additional media sources make it easier for people to gather insights from a wider range of content types.
The AI-generated audio summaries are one of the most popular features. Many users shared their summaries on social platforms, particularly because the AI voices used sound conversational, almost like a podcast. Google noticed this trend early and saw it as an opportunity to fine-tune the way Audio Overview works.
Future of AI-Driven Research
Google isn’t stopping here. There are discussions underway about adding more features, including a mobile app to make NotebookLM more accessible to smartphone users. The company also plans to introduce additional voices, languages, and control options to the audio summaries in the future.
Another idea the team is working on involves expanding the number of AI speakers in the generated audio. Right now, it’s limited to two voices, but the company has explored adding more for a multi-speaker format.
Overall, Google is positioning NotebookLM as an increasingly versatile tool, evolving from a simple research assistant to a comprehensive platform that caters to both individuals and businesses.
Open NotebookLM: A Free Rival to Google’s AI Research Tool
In a twist on Google’s AI-driven notebook, an open-source alternative has already emerged. Launched by Gabriel Chua, a data scientist from Singapore’s GovTech, Open NotebookLM challenges Google’s paid tool by offering a similar PDF-to-podcast conversion feature, but at no cost. Created using open AI models, Chua’s system was built in just a few hours, using Meta’s Llama 3.1 for language processing and MeloTTS for turning text into speech.
Unlike Google’s NotebookLM, which is embedded within Google’s services, Open NotebookLM is open for anyone to use, edit, or adapt. Hosted on platforms like GitHub and Hugging Face, the project is designed to be accessible to the broader public, including those without coding experience.
Streamlined Features for a Simple Experience
The open-source version might not have the extensive features of Google’s AI tool, but it does provide some notable options, including language support for ten languages and customization of the AI-generated voice, allowing users to pick between a more casual or formal tone. It operates on a simple interface built with Gradio, making it easy for users to convert PDFs into audio.
While Google’s NotebookLM excels with its integration into a wider array of services and deep summarization capabilities, Open NotebookLM remains focused on the core feature of creating podcasts from text, providing an alternative for users who prefer a lightweight and customizable option without the price tag.
AI Advancements and Open-Source Competition
The rapid development of tools like Open NotebookLM showcases the evolving speed at which AI solutions are being created. What once took months or years can now be accomplished in a matter of hours, thanks to available models and frameworks. However, as these AI tools become easier to build, questions arise about their reliability and the long-term impact of fast development cycles. While open-source projects allow for flexibility, they may not offer the same level of security and thorough testing as corporate-backed platforms.