“I asked AI to show me God”, “This AI filter is amazing”: Discourse, myths, and normalization of AI on TikTok

Abstract

This study explores the discourse around artificial intelligence (AI) on the social media platform TikTok. The popularity of AI has increased rapidly in recent years with the development of new AI platforms and tools such as Dall-E, Stable Diffusion, Midjourney, and Lensa AI as well as the built-in AI functionalities on TikTok. TikTok users have shown great interest in using these AI tools and sharing their experiences, resulting in various viral trends. The study aims to analyze the discourse around AI on TikTok, with a focus on how recurrent myths have influenced the public's understanding of AI and how it is increasingly normalized thanks to the playful nature that it has found in social media through trends and its embeddedness in platforms. The main focus of the study was to apply discourse analysis to the captions and comments of different videos under the hashtags #aifilter #artificialintelligence and #askingAI featuring TikTok trends. A categorizing system was defined to classify TikTok videos by whether they fell under entertainment or commentary on the subject. The corpus of videos was also then classified by defining positive and negative myths about AI found in captions and comments. The study highlights the normalization of AI through recurrent discursive patterns expressed by TikTok users, facilitated by social media platforms' integration with AI functionalities and its playful nature.