Overview
- Users can opt into the experiment through Search Labs and generate a 30- to 45-second audio summary of search results with a single tap
- The feature relies on Google’s latest Gemini models to script a conversational dialogue between two AI hosts
- An on-page audio player offers play, pause, volume controls, adjustable playback speed and direct links to the web pages cited
- Listeners can rate each overview with thumbs-up or thumbs-down feedback that Google will use to refine the experience
- Content creators warn that delivering answers in audio form could further reduce click-through traffic to original articles