Fueled by an upcoming version of Google’s Gemini, the Jarvis AI model is designed to function solely within Chrome, automating web-based tasks. It can analyze and interact with elements on the screen, such as clicking buttons and entering text after taking screenshots, although each action currently takes “a few seconds,” as reported by The Information.
This initiative aligns with similar efforts by other major AI firms to enhance task automation. For example, Microsoft is set to introduce Copilot Vision, allowing users to engage in discussions about the webpages they are viewing. Meanwhile, Apple Intelligence is expected to manage actions across multiple applications within the next year. Anthropic recently released a beta version of Claude, which has received criticism for being “cumbersome and error-prone,” and OpenAI is also said to be working on a comparable model.
However, The Information notes that Google’s plan to preview Jarvis in December is subject to change, with the possibility of a limited release for a select group of testers to identify and fix any issues.
NEWS DESK
PRESS UPDATE