The Evolution of the Automation of Automation to a MISSION of Agents

I have extensively written on the The Automation of Automation concept. I see the evolution of that concept now as let’s call them MISSIONS of Agents where there is a mission controller, many agents that execute tasks within the mission, other agents that validate the outcome of the agents executing their tasks at the highest level.

Key Agentic Evolutions Happening Now:

In 2024, Anthropic released MCP (Model Context Protocol) and then in 2025 Google has now come out with A2A (Agent-to-Agent). I guess in theory they serve different needs but to me they are a key part of the Agentic Automation Evolution.

While MCP focuses on enabling AI models to interact with tools and data and A2A focuses on enabling agents to communicate and collaborate with each other they both technically advance and enable Agent to Agent interactions. MCP just conceptually was rolled out as a tool integrator or enabler. At a high level though these tools while often platforms also are enabled quite often by some form of an Agent.

Both of them I would argue facilitate communication and collaboration between multiple AI agents. These technologies ARE enabling agents to discover each other, communicate with each other, and interface (delegate or assign) tasks to each other. This is all relatively new and brings innovations of innovations itself.

Human + AI moves to AI + Human

I previously was more focused on what I’ll call Human + AI solutions before 2025. I still saw massive concerns and issues with hallucinations enough that I preferred to create solutions that would still mostly have a Human in the Loop. With the rapid advancements in quality outcomes related to strictly scoping agents to specific manageable tasks I’m actually seeing some Agentic Systems evolve more towards AI + Human where there still is a Human in the Loop but its much much more automated and efficienct since the Human is only required to get involved if the MISSION Driven Agentic system runs into issues individual agents cannot handle.

Key Takeaways:

Key Follow Up Thoughts:

I wrote this related to Human + AI months ago.
“In this stage of AI Evolution, the use and tactic of splitting up tasks between what AI does very well and accurately with what humans do well can provide competitive advantage to business leaders. Human + AI subtask division and collaboration can improve many aspects of AI automations including:
* Quality of AI outputs.
* Quality of inputs to AI.
* Task efficiency and speed improvements without sacrificing level of quality.
* Consistency and Improved Outcomes.
“

What I took away though from Snowflake Summit 2025 and Sam Altman’s conversation though was that even he (and MOST OF US) did not see how rapidly quality and innovation could come from GenAI.

Key Agentic Evolutions Happening Now:

Human + AI moves to AI + Human

Key Takeaways:

Key Follow Up Thoughts:

What I took away though from Snowflake Summit 2025 and Sam Altman’s conversation though was that even he (and MOST OF US) did not see how rapidly quality and innovation could come from GenAI.

Leave a Reply Cancel reply