AI as Discourse Actor: A Critical Discourse Analysis of Transactional and Interactional Functions in Social Media Video Content

Abstract

The emergence of artificial intelligence (AI) within digital media has greatly changed the process of audiovisual communication on platforms such as TikTok and Instagram. In this study, transactional and interactional discourses are analyzed in relation to the construction of AI within AI-generated videos using the MCDA method. Specifically, the research explores selected AI-generated videos by focusing on the internal characteristics of the videos and analyzing key frames from each video. It was found that although transactional discourse serves the purpose of showing AI as an orderly system, interactional discourse becomes dominant by the means of multimodal elements which appeal to the emotions of viewers through their cinematic and extremely realistic quality. The use of emotional and other elements allows constructing an image of AI not only as a system but as a creative and emotionally intelligent agent as well. Thus, this study demonstrates the relationship between interactional and transactional discourse and the importance of their interplay for the construction of the image of AI within the context of its normalization. It fills the existing gap in literature which primarily focuses on textual or audience analyses rather than on the internal multimodal features of AI-generated videos themselves.