A regular protocol co -developer for communication between artificial intelligence (AI) agents has emerged. In preparation for the opportunity of an AI agent in-built its own way, a brand new standard to integrate it is required.
Cisco, Langchain, Ramaginedex, Galileo, and Glin announced on the sixth that AI agents have launched an open source community ‘AGNTCY’ to develop standard protocols that may freely exchange and communicate with one another.
The goal of the agency is to determine an industrial standard agent interoperability language in order that AI agents of assorted organizations may be freely and efficiently.
In recent months, Sales Force, Microsoft, Antroic, Open AI, and Google have been pouring out AI agents one after one other, and most are focused on certain tasks. In consequence, there are also evaluation that the ‘orchestration agent’, which integrates individual agents, will likely be the trend within the second half.
But non -Josei Panday Out Shift identified that “there isn’t a standardized infrastructure that agents can find, communicate, and work together.”
“We’ll maximize global work efficiency by constructing the Web of AI agent, just because the initial Web connects all servers and the online through the TCP/IP protocol,” he said.
The operation and management of the agency are in control of Cisco, Langchain, and Galileo, and Glin and Rama Index are participating as major contributions. If major technology firms take part in the telecommunication protocol, the structure may change.
Prior to this, Langchain has already developed a protocol for communication between AI agents. The Agent Protocol, which was released in November last yr, allows the Langchain Agent to speak with other frameworks comparable to AutoGen and Crew AI.
At the identical time, Antropic announced the model context protocol (MCP), which standardizes the tactic of modeling and AI tools connected to data sources. Nevertheless, it has not yet turn out to be an official standard within the industry.
The agency shouldn’t be just defining a typical protocol, but plans to expand right into a platform that firms can create, distribute, operate, and evaluate various AI agents through this standard.
“We’ll allow developers to search out and mix different AI agents and construct efficient workflows through the Agency platform.”
More AI firms are essential to turn out to be a worldwide standard. Nevertheless, since many firms are already developing AI agents in their very own way, it’s questionable whether agents can achieve their goals.
By Park Chan, reporter cpark@aitimes.com
