지표적 흔적에서 확률적 구성으로: 행위자네트워크 이론으로 본 AI 이미지와 사진의 존재론적 분기 = From Indexical Trace to Probabilistic Construction: The Ontological Divergence of AI Images and Photographs in Actor-Network Theory

다국어 초록 (Multilingual Abstract)

This study investigates the fundamental medial differences between AI-generated images and photographs. By employing actor-network theory (ANT), it moves beyond superficial analyses to “unblackbox” the production network of each medium. The analysis reveals photography as an “indexical network” that “captures” the world through physical causality and creates a direct trace of reality. In contrast, AI imagery is a “probabilistic network” that “constructs” images from data archives based on statistical probability originating from linguistic prompts and datasets often “inscribed” with societal biases. The study concludes that the core ontological divergence is the shift from “indexical trace” to “probabilistic construction,” and defines the AI image as a new medium: a “probabilistic generative apparatus.” Furthermore, this framework suggests that artistic value in the age of AI lies not in the final product, but in the “co-creation” process itself, which forms new relationships through human-machine interaction.