Google has launched the Model Context Protocol (MCP) Server for its Data Commons platform, enabling developers and AI systems to access public datasets using natural language. This initiative streamlines the integration of trusted data—like census and climate information—into AI training pipelines, thereby mitigating data hallucinations often arising from unverified web content. Since its introduction in 2018, the Data Commons platform aimed to consolidate diverse datasets, making them more accessible to users. The MCP Server enhances this accessibility, allowing AI agents to query data effortlessly. Google has also partnered with the ONE Campaign to develop the One Data Agent tool, which applies multiple datasets for health and finance in plain language. The MCP Server is open-source and compatible with various large language models (LLMs), available for developers globally. By empowering researchers and organizations with reliable data access, this launch marks a significant step towards improving AI reliability and functionality.
Source link
Share
Read more