Google has launched a Model Context Protocol (MCP) Server for its Data Commons platform, revolutionizing access to public datasets like census and climate data using natural language. This innovation aims to enhance the integration of high-quality, verifiable data into AI training and applications, mitigating the issue of AI hallucinations, which arise from using unverified web content. The Data Commons platform, inaugurated in 2018, has faced challenges in dataset accessibility, but the MCP Server now simplifies queries for developers and AI systems.
Additionally, Google collaborates with the ONE Campaign to develop the One Data Agent tool, leveraging the MCP Server for health and finance datasets. The MCP Server is compatible with large language models (LLMs) and is now available to developers worldwide for open-source trials, promoting improved AI reliability and enabling integration with trusted data sources. Google also encourages developers to experiment with the MCP Server through its Agent Development Kit (ADK).
Source link
