Google is preparing a revolutionary feature for Gemini that will allow the AI assistant to directly control your Android phone to perform complex tasks for you. This new capability, called "screen automation," will be able to issue commands, book rides, and interact with various applications installed on your device.
How it works
Beta version 17.4 of the Google app reveals the details of this feature, internally codenamed "bonobo." Gemini will be able to interact with certain applications on your phone autonomously, performing the necessary actions to complete the requested task. This function builds on the foundations laid by Android 16 QPR3.
However, the user remains in control: you can stop the AI agent at any time and resume the task manually. Google emphasizes that "Gemini can make mistakes" and that "you are responsible for what it does on your behalf, so monitor it closely."
Privacy questions
Google has planned several privacy safeguards:
- When Gemini interacts with an app, screenshots are reviewed by human reviewers and used to improve Google services if 'Keep Activity' is enabled
- It is recommended not to enter login or payment information in Gemini conversations
- Screen automation should not be used for emergencies or tasks involving sensitive information
Integration with 3D avatars
The update also reveals a "Likeness" feature (codename "wasabi") which refers to the 3D avatars currently used in Android XR for Google Meet calls. This integration suggests a possible use of these personalized avatars in the Gemini interface.
This feature is not yet officially launched and remains in the development stage in beta versions for now.
No comments! Be the first one.