COMMANDING IS CREATING: A SPEECH ACT ANALYSIS OF SELECTED AI PROMPTS

Abstract

This study explores the intersection of Speech Act Theory and artificial intelligence (AI) prompt generation, focusing on how user commands facilitate creative outputs in image production. As AI systems increasingly participate in content creation, understanding the dynamics of human-AI interaction becomes crucial. Through qualitative analysis of five AI prompts, the study investigates various speech acts, emphasizing directives as the primary form of user input. Each prompt is examined to identify its locutionary, illocutionary, and perlocutionary aspects, highlighting how effectively the AI responds to user requests. Findings reveal that while many prompts achieve successful outcomes, discrepancies arise when the AI fails to adhere to specific user intentions, affecting the sincerity conditions of the speech acts. The study elucidates the importance of precise prompt formulation, with the direction of fit being a key determinant of success in generating desired outputs. By employing Speech Act Theory, this research contributes to a deeper understanding of the communicative processes underlying AI interactions, offering valuable insights for creators and enhancing the effectiveness of AI-assisted content generation.