Microsoft’s Magma AI Can Manipulate and Control Robots
Microsoft just introduced Magma, a new AI model Designed to help robots see, understand and act smarter. Different from tradition AI Models, magma will process different types of data at once – Microsoft’s efforts have made a big leap towards “agent AI” Or a system that can plan and perform tasks on behalf of the user.
The model combines visual and language processing and trains video, images, robot data and interface interactions to make it more versatile than previous models.
On its GitHub page, the Microsoft research team outlines how magma performs tasks, such as how it manipulates a robot and navigates the user interface (such as clicking a button).
To develop the technology, the company collaborated with researchers at the University of Maryland, the University of Wisconsin-Madison and the University of Washington.
Launch is the development of AI agents that can automate daily life with the development of the competition for technology giants. Google has been advancing a language model focused on robotics, and Openai’s operator tools Designed to handle mundane tasks such as making reservations, ordering groceries, and filling out forms by typing, clicking, and scrolling in a dedicated browser.
