Microsoft's Vision for Collaborative AI Agents

Microsoft aims to foster collaboration among AI agents from different firms by supporting the Model Context Protocol. This protocol promises to enhance cross-company interactions and agent memory. Microsoft's forthcoming Build conference will focus on adoption standards across the tech industry, promoting an 'agentic web' akin to the internet's evolution.

Microsoft is setting its sights on a future where artificial intelligence agents from various companies can seamlessly collaborate and recall their interactions more effectively. Chief Technology Officer Kevin Scott emphasized this vision ahead of the company's Build conference, set to begin in Seattle on May 19.

During a briefing at Microsoft's Redmond headquarters, Scott highlighted the company's commitment to adopting and supporting the Model Context Protocol (MCP), a standard introduced by Anthropic. MCP aims to create an interconnected 'agentic web,' reminiscent of the early web protocols of the 1990s.

By leveraging structured retrieval augmentation, Microsoft seeks to enhance AI agents' memory capabilities. This innovative approach involves summarizing each interaction concisely to build a comprehensive dialogue roadmap, despite the significant costs associated with enhanced AI memory.

(With inputs from agencies.)