Social Sciences in China (Chinese Edition)

No. 10, 2024

Generative Governance: A New Paradigm of Governance in the Era of Large Language Models

(Abstract)

Mi Jianing

“Generative governance,” an emerging paradigm of public governance, is based on generative artificial intelligence technology. It systematically empowers decisionmaking and execution optimization in public affairs governance through comprehensive data perception, crossdomain knowledge extraction, and multi-scenario strategy generation. The rise of generative governance is jointly driven by the support of ubiquitous data, large language model algorithms, and highperformance computing. Theoretically, the efficacy of generative governance depends on the generative “emergence” of knowledge, selforganization and evolution, and the interplay between instrumental rationality and value rationality. In practice, through data feeding, algorithm iteration, and model finetuning, generative large language models can propel transition from specialized technical governance to general public affairs governance, achieving multi-scale governance from micro-level individuals to macro-level systems in a wide, complex, and constantly changing public affairs environment. Generative governance will herald a new era of more intelligent, efficient, and humane public governance.