China Social Science Review

No.4, 2023

 

A Review of the Technopolitics of Large Language Model

(Abstract)

 

Wu Guanjun

 

In the ranks of knowledge producers, a stranger to the eyes of mankind emerged: Large Language Model (LLM) as generative artificial intelligence. The LLM, represented by ChatGPT, demonstrates a near-general-purpose level of intelligence, while at the same time a large number of users report a variety of errors and even frequent errors. It therefore leads to a systematic inquiry into the intelligent form of the LLM from a techno-political perspective, based on which the causes of its generative errors and ideological biases are analyzed. The U.S. government requires that “American values” must govern the development and application of artificial intelligence systems, which means that the crucial thesis in the age of artificial intelligence remains critical analysis of ideology.