Microsoft unveiled magma: a new AI model to control software and robot

Microsoft Research has introduced a new AI model called Magma, which can mark a large progress in artificial intelligence to control both software interface and robotic system.

Listen to the story

Advertisement
Microsoft unveiled magma: a new AI model to control software and robot

Microsoft Research has introduced a new AI model called Magma, which can mark a large progress in artificial intelligence to control both software interface and robotic system. Magma combines visual and language processing, allowing it to work in both digital and physical world, making it a potentially versatile AI model.

Unlike many existing multimodal AI systems, which rely on different models to explain data and perform tasks, Magma integrates these abilities in a system. Microsoft claims that it makes magma unique, as it can process data like data, images and videos and work originally on it, whether it is navigating the software or controlling robot yes. This advancement can lead to more autonomous and intelligent AI systems that are capable of operating in various scenarios.

Advertisement

Magma’s development has been a collaborative effort between Microsoft and major educational institutions, including Qest, University of Maryland, Visconsin-Madison University and Washington University. The aim of AI is only to go beyond answering questions or execute a single command, as Microsoft considers it as a step towards creating an agent AI system. This means that AI can plan autonomously and perform multistap tasks to achieve complex goals without human intervention.

In his research, Microsoft highlighted how Magma can craft plans based on a described target and take action to fulfill that purpose. By taking advantage of the available visual and language data, Magma can handle complex tasks in both virtual and physical settings, with a wide range of applications in industries such as manufacturing, healthcare and digital automation.

Other technical companies like Openai and Google are also discovering the ability of agents AI. Openai uses Openai with projects such as operators focus on performing tasks in web browsers, while Google is developing agent AI with its Gemini 2.0 initiative. However, the one who makes the magma isolated is its unified approach to perception and action, possibly gives it an edge into real -world applications.

LEAVE A REPLY

Please enter your comment!
Please enter your name here