Post by Ritikesh Choube

LLMs Are Getting Smarter Day by Day But Are They Truly Aware of the System They Live In? As AI researchers, we've made great progress in giving large language models (LLMs) context for example, feeding them previous chats or documents so they understand the task better. But now comes the next big challenge: How can an LLM understand and operate within the context of the entire operating system (os)? Think about it your apps follow rules to interact with your OS - saving files, opening browsers, accessing databases, etc. So why not give AI agents a similar rulebook? That’s exactly what’s happening right now in the AI world: A2A (Agent-to-Agent Protocol)- by Google https://lnkd.in/dw9BGe4J MCP (Model Context Protocol)- by Anthropic https://lnkd.in/df28rptJ These are the protocols which provides a structured way for AI Agents and LLMs to understand system level context, access tools, and collaborate effectively They are like pipelines or playbooks that LLMs and AI agents can follow to: # Interact with other agents # Access tools and databases # Understand local system context # Work together, like real teammates Think of it as teaching LLMs how to talk, " not just with users, but with the entire digital ecosystem around them. #llm #AgenticAi #MCP #A2A #Ai