Google: AI Now Reads Business Contracts Aloud as Podcasts
Google Drive is getting a new AI feature that automatically converts important PDF documents into podcast-style audio summaries.
(Image: Stock-Asso/Shutterstock.com)
Google is expanding its cloud storage service Drive with an AI-powered feature that automatically converts PDF documents into audio summaries. Users can generate an audio file in podcast style from extensive documents such as industry reports, contracts, or meeting minutes with a single click.
The new Gemini feature is based on the same technology used in Google's note-taking tool NotebookLM. It includes, among other things, automatically generated audio discussions between two AI voices. The generated audio files last between two and ten minutes, depending on the scope of the source document, and are automatically saved in a dedicated "Audio Overviews" folder in the user's Google Drive.
After creation on a desktop, users receive an email notification once the audio file is complete. The summaries can then be played back from any device with access to Google Drive, including mobile devices. Google primarily positions the feature for users who want to "read" long documents while engaged in other activities, such as commuting or exercising.
Unlike simple text-to-speech output, the AI summarizes the essential content of the PDF and presents it in a dialogue format. To do this, the AI technology analyzes the document content and extracts the key messages before converting them into a as natural-sounding as possible audio discussion.
Videos by heise
Significant Limitations at Launch
At market launch, the feature exclusively supports English-language PDF documents. Google has not provided any information on when further languages will be supported. Other file formats, such as Word documents or PowerPoint presentations, will also not be supported initially.
The Audio Overviews are part of the Gemini offering for Google Workspace and are therefore not available to all Drive users. The feature is being rolled out to users via both the Rapid Release and Scheduled Release channels. Workspace administrators can configure the feature for their organization. Information on this can be found in the Workspace Blog.
(fo)
