AI and AR are on track to move forward seamlessly, a vision already mastered by Meta and recently highlighted at Google I/O with the introduction of Project Astra. This new AI initiative by Google promises continuous scanning of camera feeds to provide contextual understanding of the surroundings, positioning it as a blueprint for future AI-powered glasses and devices.
Google made Astra “the future of AI assistants,” focusing to create a universal AI agent that can assist in everyday life with speed and conversational symphony. Unlike current AI gadgets like Meta’s Ray-Ban glasses, Rabbit R1, and Humane AI Pin, which suffer from slow response times, Google’s Astra demo showcased near-instant responses comparing to others. The demo involved a phone’s camera scanning a room in real-time, demonstrating the agent’s rapid contextual recognition.
In a particularly satisfying part of the demonstration, the AI was shown running on a pair of glasses, showing Google’s commitment to wearable AI technology. The person using Astra switched from a phone to camera-enabled glasses, asking questions like “Where did I leave my glasses?” and receiving prompt, accurate responses such as “On the corner of the desk next to a red apple.” These glasses resembled the AR translation glasses Google announced but never released two years ago, and mark a significant evolution from Google Glass, launched 11 years ago.
Google is already collaborating with both Samsung and Qualcomm to develop a mixed reality platform, hinting at potential camera-enabled AI glasses. This direction aligns with Meta’s vision, as stated by Mark Zuckerberg and Andrew Bosworth, that AI is the key to advancing XR devices.
Google DeepMind’s Project Astra aims to build a universal AI agent capable of real-time reasoning and lightning-speed response. The demo showcased Astra identifying objects and locations almost closely, based on Gemini 1.5 Pro mode and other task-specific models. The technology continuously encodes video frames to minimize response times.
Furthermore, Google showed the output Gemini Live on smart glasses, overlaying information directly onto the user’s vision. These glasses appeared to be an iteration of the translation glasses prototype shown at I/O 2022, illustrating Google’s strides in integrating AI with AR for practical, everyday use.
