Why It Matters
Stateful protocols like MCP are essential for reliable, long‑running AI interactions, positioning enterprises to deploy resilient conversational agents at scale.
Key Takeaways
- •MCP adds statefulness to AI communication protocols
- •Enables conversation resumption after interruptions
- •Supports elicitation and sampling for richer AI interactions
- •SEP-1391 will introduce new MCP specification enhancements
- •Differentiates from gRPC and HTTP via built‑in fault tolerance
Pulse Analysis
The rise of autonomous AI agents has exposed a limitation in traditional stateless APIs such as gRPC and HTTP. While these protocols excel at request‑response patterns, they lack the memory needed for prolonged, context‑aware dialogues. MCP—short for Multi‑Channel Protocol—addresses this gap by embedding state directly into the communication layer, allowing servers to track conversation history and resume interactions after network failures. This design mirrors human phone calls, where a dropped line does not erase the discussion, and it lays the groundwork for truly persistent AI assistants.
Beyond simple state retention, MCP introduces two under‑utilized capabilities: elicitation and sampling. Elicitation lets the server pose clarifying questions to users, turning a one‑way prompt into a dynamic exchange that improves answer relevance. Sampling empowers the protocol to invoke the underlying large language model on‑the‑fly, enabling adaptive response generation without external orchestration. Together, these features create a fault‑tolerant, self‑healing communication fabric that can survive interruptions while maintaining conversational continuity—something stateless protocols struggle to achieve. This approach also reduces latency by avoiding round‑trip orchestration.
The industry is already responding. Prosus’ MLOps Community has backed the upcoming Specification Enhancement Proposal 1391, which refines MCP’s error‑handling semantics and expands its tool‑integration model. As AI agents move from experimental labs into production workloads, a stateful protocol becomes a strategic asset for enterprises seeking reliable, scalable interactions. The talk at the Computer History Museum, led by the creator of Apache Mesos and a CNCF veteran, signals that MCP is poised to become a cornerstone of next‑generation AI infrastructure, influencing both open‑source roadmaps and commercial offerings.
Original Description
March 3rd, Computer History Museum CODING AGENTS CONFERENCE, come join us while there are still tickets left.
Thanks to @ProsusGroup for collaborating on the Agents in Production Virtual Conference 2025.
Recently, we’ve heard smart people we respect say things like: “There's a lot of buzz around MCP. I'm not convinced it needs to exist.” In this talk, we will argue that MCP differs from other modern protocols like gRPC and HTTP primarily due to its inherent statefulness.
And, that this statefulness is required to bring about the rich, intelligent, long-lived, human-like communication that we want with AI. It is a feature, not a bug.
Consider a human-to-human phone conversation: you're on the phone, the call drops, and you dial back. You don’t expect to begin your conversation all over again. Instead, you'd anticipate resuming from where you left off, or at least very close to it. To achieve something similar from human-to-computer, the protocol needs ways to pick up where the chat left off.
More so than schemas and a clean separation of prompts, resources, and tools, this is what excites us the most about MCP. This talk will explore the aspects of MCP that enable fault-tolerance as well as features that have remained relatively obscure like elicitation (which enables the MCP server to ask questions or elicit feedback from humans), and sampling (which enables the MCP server to invoke the LLM itself).
We will also discuss the current MCP client landscape as well as some important SEPs (Specification Enhancement Proposals) coming down the pike, like SEP-1391, in which we have been involved.
Bio //
Creator of Apache Mesos. Co-founder of Mesosphere/D2iQ and reboot.dev. Early Twitter engineer. Founding member of the Technical Oversight Committee for the CNCF.
A Prosus | MLOps Community Production
0
Comments
Want to join the conversation?
Loading comments...