Google is rapidly integrating AI-generated podcasts into its offerings, making it a noteworthy addition for users. This feature allows the creation of two-person AI podcasts derived from Deep Research reports and even enables the generation of a personalized news recap through a Daily Listen podcast. Plans are underway to incorporate this functionality into Google Search as well.

Currently available as an experimental feature via Google Labs, this tool generates a condensed, five-minute AI podcast based on your search inquiries.

Activating Audio Overviews in Google Search

Credit: Jake Peterson

To activate this feature, navigate to your Google Labs page and locate the Audio Overviews section. Depending on your region, you will either be able to turn on this feature directly or opt to join a waitlist. It’s important to note that this global roll-out will not be instant, as is typical with many new AI functionalities from Google.

Understanding Audio Overviews in Google Search

The concept of this new search feature is derived largely from Gemini, which has its roots in NotebookLM. Named Audio Overviews, this functionality initially allowed users to create a 10-minute AI podcast episode on a variety of subjects; however, the latest iteration comes with certain constraints.

Once this feature is activated in Google Search, users will notice a prompt labeled “Generate Audio Overview” as they browse through relevant search results. Currently, the criteria for which results qualify remain somewhat ambiguous. Simple inquiries like “what are some nearby cafés?” will not trigger this feature, and more complex subjects—such as analyzing investment trends across Asia—may be better suited for specialized tools like Deep Research.

Audio Overviews are designed to assist with queries that fall in between the two extremes. For example, if one wishes to gather quick insights about a character from Lord of the Rings or get recommendations for essential Japanese kitchen knives, a relevant Google search followed by clicking the “Generate Audio Overview” button will activate the feature. Users can expect to receive their audio overview after a brief waiting period of 30 to 40 seconds, which is significantly quicker than the 2 to 5 minutes typical of Gemini, yielding a concise five-minute summary.

The audio player for this AI-generated podcast will remain accessible as users browse the results, complete with links to the sources used. If any inaccuracies are detected, there’s an option to provide feedback by giving it a Thumbs Down. However, it’s crucial to recognize that AI tools like these rely on Large Language Models, which can occasionally produce erroneous information. Always verify the sources linked within the Audio Overviews feature before sharing the content elsewhere.