China's Central Cyberspace Affairs Commission has issued the 15th Five-Year Plan for national informatization, published with an accompanying Q&A on the Cyberspace Administration's site on 23 October. According to the Q&A, the plan targets 2030: major progress in informatization, a fully built next-generation communications network, a commercial 6G launch at the appropriate time, a basic national blockchain network, and faster layout of what the document calls an intelligent internet. The plan sets 12 headline indicators, covering internet penetration, 5G and 5G-A base stations per 10,000 people, IPv6 traffic share, power usage effectiveness for newly built large and very large computing facilities, R&D intensity in computer and electronics manufacturing, valid domestic invention patents in core digital-economy industries, the penetration rate of new-generation intelligent terminals and agents, the digital-economy core sector's share of GDP, the share of firms with fully digitized key business processes, cross-border e-commerce volume, advanced digital literacy among adults, and electronic social-security card coverage. The Q&A names the indicators but does not give their target values. Work is organized as three systems, five levels of improvement and two supporting foundations: 10 areas, 25 tasks and 7 special columns, including projects for information-infrastructure innovation and cybersecurity science and technology, an action for intelligent social development and governance, and one for facilitating cross-border data flows. The plan also calls for security assessment of new technologies and applications and stronger data-security and personal-information protection.
Our takeThe interesting part is what gets counted. Making intelligent-terminal and agent penetration a national indicator moves the measurement target away from model benchmarks and toward deployment breadth, which is a different thing to optimize and a harder one to define. The PUE indicator for new large computing facilities ties capacity growth to efficiency at the point of construction rather than operation. So what: the operational definition of agent penetration will matter more than the number attached to it, because whatever counts as a deployment becomes what gets built toward.