Google has introduced its latest AI-powered feature called ‘Audio Overview’ through Search Labs, generating audio summaries of search results in the style of two AI-generated podcast hosts. This feature is currently available as an opt-in in Search Labs, limited to the U.S. and English summaries.
How Audio Overview Works
To access Audio Overview, users must first enable it in Search Labs. Once activated, certain Google searches will display an Audio Overview box, typically below the regular AI Overview and ‘People also ask’ sections. Users can generate an audio summary by tapping a button, with the processing time varying depending on the search query.
Characteristics of Audio Overview
- The audio clip is generated on the spot and can vary in content if regenerated.
- Summaries range from 3 to 5 minutes in length.
- Features two AI-generated voices discussing the search topic in a podcast style.
- Voices are more lifelike than typical robotic voices, with tone changes and conversational language.
- Despite advancements, the voices lack natural rapport and sometimes sound like they’re reading from a teleprompter.
Implications and Concerns
While Audio Overview shows potential, particularly for those with vision impairments or relying on screen reading tools, it also raises concerns:
- The risk of spreading misinformation through AI-generated content.
- Potential threat to jobs such as voice acting due to advanced AI voice technology.
- The AI voices, while realistic, can mispronounce words and sometimes sound awkward or stiff.
User Transparency and Verification
Google provides transparency by showing which search results were used to generate the audio summary, allowing users to verify the information. This feature is a double-edged sword, offering helpful accessibility while posing risks associated with AI-generated content.
The balance between innovation and responsibility is crucial as AI technology continues to evolve.
