Generative AI software like ChatGPT, Gemini, Claude, and others offer plenty of similar features. There’s a lot of overlap, and it’s relatively easy to switch from one AI product to another or use multiple options at the same time. These AI programs also get increasingly more complex with each new update. The companies behind them try to both outflank the competition and match brand-new features that might be available from rivals.
Google just did that on Tuesday, announcing a new Canvas feature for Gemini that will work exactly like the ChatGPT Canvas. Previously, OpenAI launched a Deep Research feature that Gemini got before ChatGPT.
But Google’s Tuesday AI announcement is also exciting for a different reason. Canvas will work just like ChatGPT, so there aren’t any surprises there. What’s exciting is Google’s decision to expand one of its lesser-known Gemini features that ChatGPT doesn’t have.
I’m talking about the Audio Overviews feature that launched last September for NotebookLM. Audio Overview lets you turn any documents you upload to the AI into podcast-like shows. It creates AI characters that discuss the files you’ve just uploaded, and it’s a feature I can’t wait to play with.
Summarizing larger files is one of the things I do all the time with ChatGPT. It’s part of the way I work, and it’s an amazing tool to have on hand. The more you tweak it, the better the results you get.
Audio Overview in NotebookLM adds another layer of complexity and varoety to the type of summaries you can get with AI. Audio Overview makes them interesting, with two AI hosts engaging in a conversation that covers the file(s) you upload to NotebookLM.
The podcast style of summary isn’t necessary for all types of documents, but it can certainly offer a deeper dive and make a better impression. Rather than reading the summary, you have the AI talk about it with itself.
You could argue that ChatGPT might be able to do something similar with Advanced Voice Mode. You can use it to read to you the summaries ChatGPT has created for you. But it’s not quite like Audio Overview.
The fact that Audio Overview was buried in NotebookLM until Tuesday might not have been the best idea for Google. Tech-savvy AI users might know about it, but Audio Overview wasn’t immediately obvious to new users.
Starting Tuesday, the feature is moving to Gemini. Audio Overview will work with all sorts of content, including documents and files you upload but also Deep Research reports that Gemini can create for you. As a reminder, Google updated Deep Research, having the Gemini 2.0 Flash model power the feature a few days ago.
Audio Overviews in Gemini will work just like in NotebookLM. The AI will create a podcast that covers the topic at hand:
Gemini will create a podcast style discussion between two AI hosts who, with just a click, launch into a lively deep-dive conversation based on your uploaded files. They’ll summarize the material, draw connections between topics, engage in a dynamic back-and-forth and provide unique perspectives.
Audio Overview will be rolling out to Gemini and Gemini Advanced subscribers globally on Tuesday. It’ll be available on the web and in the Gemini app in English initially, with more languages to follow.
To get started with Audio Overview, you’ll need to upload the files you want a podcast on and then click the suggestion chip that appears above the composer (prompt) bar. Alternatively, look for Audio Overview in the triple-dot menu in the Gemini mobile app, as seen in the example above.