튜링 테스트를 통한 챗GPT의 거절 화행 비교 연구 -비윤리적 질문을 중심으로-
A study on verification of refusal speech acts of ChatGPT through the Turing test -Focusing on unethical questions
- 발행
- 2023
- 소속·발행
- 한양대학교
- 출처
- 국내 KCI
- DOI
- 10.22832/txtlng.2023.55..002
- 원문등록
- 2023-12-31
- 원문
- 원문 보기 ↗
개념
키워드
Unethical Questions, Refusal Speech Acts, Refusal Strategy, ChatGPT, Generative AI, Turing Test, Large Language Model, ChatGPT's level of Humanization, Unethical Questions, Refusal Speech Acts, Refusal Strategy, ChatGPT, Generative AI, Turing Test, Large Language Model, ChatGPT's level of Humanization
초록
The present paper aimed to examine the patterns of refusal speech acts in response to unethical questions from both humans and ChatGPT and strategically assess whether ChatGPT's responses reached a level of similarity to human responses. To conduct the research, refusal speech act strategies of humans and ChatGPT were analyzed and compared in the context of unethical questions. The research findings revealed differences in the refusal speech act strategies of humans and ChatGPT, and these differences were found to diminish ChatGPT's level of humanization.