By buld_online 
Google is testing a new feature that turns complex search queries into short, AI-generated podcast-style conversations. Called Audio Overviews, the experimental tool uses the company’s Gemini AI models to create dynamic, two-voice summaries that sound like a mini podcast, offering a more engaging, hands-free way to absorb information.
The feature is being rolled out through Google’s Search Labs and appears for more nuanced questions, not basic queries like the weather or calendar dates, but broader topics such as the significance of Flag Day or the use of Argon in science. When available, users will see a “Generate Audio Overview” button under the search bar. After a brief wait, Google generates a back-and-forth conversation between two AI voices, summarising information pulled from top search results, all of which are cited.
Audio Overviews are not entirely new. They first appeared in Google’s AI note-taking app, NotebookLM, and were later added to the Gemini app with additional features like follow-up questions and a “Deep Dive” mode. The Search version, while simpler, still allows users to adjust playback speed (from 0.25x to 2x) and volume via a built-in media player.
For now, the feature is available only to users who opt in via Search Labs, and access appears to be limited, possibly to those using the Google app. Google is positioning the tool as a more natural way to learn about unfamiliar topics while on the go.