Google Chrome will integrate Gemini on May 21, enhancing browsing for Google AI Ultra and AI Pro subscribers in the U.S. Users must have the latest version of Chrome with English as their preferred language. Gemini will assist users with tasks and navigate the web with minimal input, allowing commands like directing to specific sections of a website. Future updates will enable Gemini to work across multiple devices. Google aims for Gemini to execute tasks autonomously and develop it into a "World Model" by combining Project Mariner's multitasking with Project Astra's visualization capabilities.