On Oct. 28 the National Data Administration convened a meeting of industry
high-quality dataset chain-master units, instructing them to map artificial
intelligence data needs and accelerate dataset development. Chain-master units
were told to base work on their own needs, proactively identify upstream and
downstream external demand, and follow a demand-led approach for precise dataset
construction. They were directed to pilot benefit-sharing and commercial-return
mechanisms, compile industry data maps, promote distributed aggregation of
industry data and register datasets with the national dataset management service
system. The administration also ordered establishment of application-driven data
feedback loops to enable data return and iterative optimization, continuous user
feedback and dynamic dataset updates, and stronger industry-chain coordination
to broaden cooperation and build ecosystem capacity.
