@jam.with.ai: Protocols got weird 😅 Note: I know not all of them are the protocols. I’m just projecting that protocols with respects to LLMs are quite different and very loose - and not like how the earlier protocols were. With LLMs things are more loosely stitched because of its probabilistic nature. It’s just an interpretation. TCP/IP, DHCP, HTTP, and TLS standardize how systems connect, exchange data, assign addresses, and secure communication. AI protocols operate at a higher layer. 1. MCP standardizes how models connect to tools, APIs, and data sources. 2. A2A standardizes how agents discover each other, delegate tasks, share status, and return results. 3. SKILLS.md defines reusable instructions for how an agent should perform specific work. 4. LLMS.txt gives AI systems a structured summary of useful website content. That is why the comparison is funny. Older protocols feel like strict engineering standards. Newer AI protocols often look like JSON, Markdown, metadata files, and documentation that everyone agrees to interpret the same way. But the goal is still interoperability. Traditional protocols standardize data transfer. AI protocols standardize context, capabilities, tools, and coordination. The real challenge is not inventing another format. It is making sure different agents interpret it consistently, respect permissions, handle failures, and remain compatible as implementations change. What do you think about it? #ai #softwaredeveloper #machinelearning

Shirin

Open In TikTok:

Region: DE

Saturday 18 July 2026 19:06:02 GMT