Page 159 - 《国际安全研究》2023年第3期

Vo1. 41, No. 3, May/June 2023

Abstract

3 Subject Dispersion and the End of Subject Responsibility˖Possible
Impact of ChatGPT on Global Security Practices

GAO Qiqi and ZHANG Yingwen

[Abstract] In the era of artificial intelligence, the general AI model represented by
ChatGPT may impact the “subject responsibility system” in global security interactions,
thus giving rise to a series of problems such as the fuzzification and lack of
responsibility of security subjects, the discretization of security object responsibility
and the difficulties in identifying accountability for security issues. Given that the
technological development of artificial intelligence technology is featured by iterative
evolution from being weak to being strong, the forms and degrees vary greatly
concerning the “subjectivity crisis” and the “responsibility crisis” at different
technological stages. In the era of weak artificial intelligence, the empowering effects
of the low intelligence general large model may strengthen the responsibility of the
marginal security subjects in the field of security interaction, where security
interaction presents a mode featuring coexistence of multiple security subjects and the
discretization of security object responsibility. In the era of strong artificial
intelligence, the general model is designed as a “humanoid” subject, which is a new
one among the existing multiple security subjects. The security interaction has taken
on irregular forms. The existing security subjects may show the characteristics of
over-reliance on algorithm and the erosion of subject power, while security objects
may display the general features of shirking their responsibilities and the end of
subject responsibility. In order to solve the above problems, we need to start at the
“intersubjectivity” level of security subject interaction. In the era of weak artificial
intelligence, blockchain and artificial intelligence technology can be utilized to form
synergy and refine prior responsibilities. In the era of strong artificial intelligence, it
is advisable to forge a “guardian” system based on coordination among major
countries and the establishment of the “semi-subject” identity of intelligent agents so
that global security practices can be carried out along the logic line of intersubjectivity.
[Keywords] Artificial Intelligence, ChatGPT, security practices, subjectivity,
responsibility
[Authors] GAO Qiqi, Professor, Institute of Political Science, East China University
of Political Science and Law; ZHANG Yingwen, Ph.D. Candidate, School of
Government, East China University of Political Science and Law (Shanghai, 200042).

28 The International Community’s Multiple Perceptions of GSI and
China’s Promotion Strategies
WANG Mingguo

[Abstract] The Global Security Initiative (GSI) is a Chinese solution to global security
dilemmas. The GSI has received enormous attention from the international community

· 157 ·