Exploring the Future of Visual Interaction with Gemini Live Astra’s World-Facing Camera [Video]
Overview of Gemini Live Features
Gemini Live, powered by Astra, is expanding its visual functions on Android devices. This article provides a hands-on experience and highlights the key features, including world-facing camera usage and screen sharing capabilities.
Accessing Gemini Live Astra
- Subscription Requirement: Requires Google One AI Advanced subscription. Free trials are available for Pixel 9 Pro (12 months) and Galaxy S25 models (6 months).
- Launch via Gemini App: Launch the Gemini app to access new visual modes, which include camera and screen sharing controls.
- Voice Activation: Use the “Hey, Google” wake phrase or the 'Share screen with Live' button for AI-powered screen sharing.
User Experience
- Camera Functionality: The camera interface is user-friendly and familiar, resembling the Pixel camera UI. It supports both rear and front cameras but lacks zoom capabilities.
- Focus Issues: Focus can be finicky, especially for close-up subjects, requiring users to keep still or ensure the subject remains stationary.
- Use Cases: Best suited for deeper queries like object identification, food information, and real-time translations. On-screen text is not provided; responses are audio-based.
Screen Sharing
- Current Limitations: Shares the entire screen rather than a single app. Visual feedback is limited to audio cues and information.
- Potential Improvements: Future integrations may allow for adding items to shopping lists or calendar events directly from shared screens.
Accuracy and Considerations
- AI Limitations: Gemini can provide incorrect information, especially with less familiar objects. Clear views within the viewfinder help mitigate errors.
- Trust Issues: Users should not blindly trust AI information due to potential hallucinations and inaccuracies.
Future Prospects
- AR Integration: The current features are a step towards Google’s AR platform vision. While limited, they show promise for improvement and better integration with other services over time.