A preliminary study on the empowerment of generative artificial intelligence in think tank research

Generative artificial intelligence (GAI) possesses advantages in data processing, content generation, simulation exercises, and other aspects, and has been widely applied in various fields. It also provides new possibilities for reshaping knowledge production and decision support models in the field of think tanks. Based on the current achievements and experiences of AI-empowered think tank research, this paper preliminarily explored the application practices and challenges faced by GAI in think tank research. The study found that human-machine collaboration was a key mode of GAI empowering think tank research. Relying on the expertise of experts and the technological advantages of GAI, it could empower research processes such as data processing, topic selection planning, and output generation, improving the efficiency of underlying data processing and promoting the production and dissemination of research results. In addition, GAI might also bring about issues related to technological dependence and subjectivity, authenticity and bias, as well as cost and process reengineering. This paper proposes relevant insights and suggestions for these issues.