Data Commons MCP Server

Google just announced the Data Commons MCP Server. This is huge for enriching the usefulness of agents and LLMs. And it points to a larger movement towards agentic data providers.

My vision is that there will be a huge number of data resources available for agents to draw from in the relatively near future. It’s a logical progression of the movement towards agentification of the internet and the economic and product drivers that that implies. And interestingly, I think these things will become a real driver of the old-school dream of a semantic, machine-operable web.

There are other interesting movements in the space. Google recently introduced the Agent Payments Protocol (AP2), enabling agents to autonomously pay for resources (with merchants that accept the protocol). Among other things, this would enable agents to access paid data resources. That could include raw data sources, curated single-source data, or pipelines that draw in different sources to provide advanced analytics and insights.

On the entrepreneurial side there are clearly opportunities here to establish agent-centric data brokers, if you have some insight into data sources and analytics that companies would benefit from having ready access to.

While it may have made sense to publish things like this in the past as APIs, there’s a certain overhead involved in consuming and sense-making at that level. Through protocols like MCP and AP2, agents will be able to move more quickly to discover, consume, and make use of rich data sources.

Whether the AP2 protocol or even MCP crystallize into widely adopted standards, the movement in this direction is clear.

What data do you wish you had at your fingertips?

References:

• Introducing the Data Commons Model Context Protocol (MCP) Server: Streamlining Public Data Access for AI Developers
• Powering AI commerce with the new Agent Payments Protocol (AP2)