I was looking at MCP from Anthropic Introduction - Model Context Protocol . While going through it I got few questions in mind which I though to take opinion from the community.
- First of all, I realize that we need a standardization, but that standardization needs to also include inter agent communication protocol. Is there a need to create an open standard for agentic systems encompassing every component of the architecture
- Should these standards be created by product companies like Anthropic or a body who will not be biased to any products or platforms
- Should we also think of ensuring that whenever we create an agentic system, by default there should be a capability to switch off the agent whenever needed.
On the 3rd point, the question comes…If the agent becomes sentient in future, will it be morally right to switch of the agent.