Photo source: Google
Google has just introduced an innovative feature in its NotebookLM application that allows users to create AI-generated podcasts based on their notes.
For those unfamiliar, NotebookLM serves as a personalised AI research assistant, which uses Google’s Gemini 1.5 Pro model. Users can upload a variety of document types, such as Google Docs, PDFs, Google Slides, and web URLs. The app then quickly becomes knowledgeable about the content as it offers source-based responses complete with in-line citations.
The Audio Overview feature, first introduced at Google I/O as “Illuminate,” allows users to generate audio conversations between two AI hosts with a single click. These hosts summarise and discuss the uploaded documents, focusing on key points, making connections between topics, and presenting information in a lively, conversational style.
This new feature builds upon NotebookLM’s existing functionalities, which already assist users in interacting with their notes, transcripts, and research documents.
Additional features of NotebookLM include:
Google has clarified that the AI-generated discussions are not intended to provide a thorough or unbiased perspective on a topic but are merely a reflection of the user’s notes. While promising, there are also limitations to the Audio Overview feature, including a potentially lengthy generation time for the audio discussions and its current availability only in English. As with many AI tools, accuracy can vary.
To try the feature, users can open an existing notebook in NotebookLM, navigate to the Notebook guide, and click the “generate” button to create an audio conversation. This allows users to experience how their research can be transformed into an engaging audio format.
