인공지능 기술 신화로서 의인화 비판: 챗봇 ‘이루다’ 사례 연구

This study considers AI anthropomorphism as a powerful component and mechanism of today’s AI technomyth. This paper critically analyzes the chatbot Iruda as a case from a constructivist perspective that reveals the ways in which this technological myth has been created and reproduced. Through the critical analysis of the Iruda case, this study observes that the anthropomorphism of AI tends to reproduce and reinforce the techno-neutralism and tech-fetishism of AI by separating the dual aspects of its technological engineering and technological culture. However, the anthropomorphisation of AI cannot be reduced to the strategic intentions of developers or specific actors, but, is a network effect that involves multiple actors (or actants) within the social debate. Today, the logic of intelligent information technology has taken on the appearance of independent and autonomous artificial objects in close relation to the myth of ‘human-like AI’. Specifically, the chatbot technology used in Iruda, despite being co-constituted by a hybrid network of AI, the technology itself, developers, government regulators, and society (anonymous AI users), paradoxically contributes to the techno-myth of AI sublime by concealing its relational aspects through the logic of “human-like AI” (anthropomorphism). This study intends to refute the replacement in which the problematic nature of the socially multilayered entanglement of AI is reduced to the problem of solving technological defects or flaws in a narrow sense. For doing that, this study critically explores the anthropomorphic processes of AI that are mobilized to reinforce technological determinism: persona (the first anthropomorphism), sexual objectification (the second anthropomorphism), childlike-ness (the third anthropomorphism), and finally the Personal Information Protection Commissioner’s policy intervention into the case of Iruda.