Google has launched a new version of its Gemini AI model, enabling advanced “agents” to perform tasks in digital and real-world environments. Project Mariner, a browser-based assistant, navigates websites and handles actions like filling shopping carts. Project Astra, initially embedded in smart glasses, offers conversational AI capable of retrieving and remembering user data.
Sundar Pichai, Google’s CEO, highlights Gemini’s improved task execution and external tool usage. However, Google promises human oversight for significant actions. As competitors like Amazon and Cohere also develop AI agents, Google aims to set benchmarks for responsible use.
Want to know more? Check out the source code on The Logic.