How to use NotebookLM’s new tools for better reporting and research

NotebookLM's latest upgrades bring deeper insights, better collaboration, and hands-free brainstorming for media professionals.

Credit: Midjourney, modified by The Media Copilot

If you’ve been following this column, you know I’m a big fan of Google’s NotebookLM, its AI-enhanced research organizer and assistant. Its ability to parse dozens of notes and other documents makes it invaluable for reporters, media professionals, and information packrats like me. (I tend to keep everything when I’m doing interviews.)

But there was always something gimmicky about the audio overviews — the feature that put NotebookLM on the map last fall but seemed more like a party trick than a useful tool. Google recently added an interactive mode, which essentially lets you act like a call-in listener to the two “hosts” and ask questions, something journalists tend to be pretty good at.

This, along with several other improvements, led me to take another look at NotebookLM and see what else it offers inky wretches and the PR people who love/hate them.

Get up to speed on NotebookLM

Before we dive into NotebookLM’s advanced features and what they can do for journalists specifically, a quick recap: NotebookLM lets you create your own personal knowledge base out of uploaded documents, videos and audio files. It supports PDFs, text, markdown, and audio files, and it allows users to access Google Drive for Docs and Slides. It can also extract text from website URLs and generate transcripts from YouTube videos. Additionally, you can paste text into a field to input data.

It can summarize news from YouTube URLs or newspaper articles, find quotes in original sources to verify their contents and context and even identify inconsistencies in notes for follow-up fact-checking or suggest potential angles for a follow-up story. Every chat response can be saved as a note for future reference.

Finally, the audio overviews feature mines all your source data to generate a podcast-style narrative between two synthetic hosts. When I wrote about it back in October, I thought it might be useful for commuters and auditory learners, but it lacked customization and professional polish.