Microsoft Research has made a significant advancement in the field of artificial intelligence by open-sourcing Agensh, a multi-agent harness designed to operate without a central orchestrator. This framework enables up to 1,024 agents to function collaboratively, sharing resources such as a Git repository, chat functions, and a findings board.
Agensh is particularly notable for its scalability, allowing users to adjust the number of agents from a single unit up to 1,024 within a specified budget of six hours on ProgramBench. This flexibility could lead to enhanced productivity and innovation in various AI applications, as teams can organize and manage multiple agents autonomously.
The release of Agensh highlights Microsoft’s commitment to fostering open-source solutions that empower developers and researchers to explore new frontiers in AI. By facilitating self-organization among agents, this tool promises to streamline workflows and improve collaborative efforts in AI-driven projects.
Compiled automatically by the Tech AI Newsdesk from public AI-news sources and summarised in our own words.
