Perception and Evaluation of AI Generated Art

Abstract

Background: This research navigates the evolving landscape of human-AI interaction, exploring key dimensions of perception in the burgeoning era of artificial intelligence (AI). Delving into trust, ethical concerns, industrial impact beliefs, and image complexity, the study seeks to unravel how individuals attribute creative authorship to AI-generated content. Methods: Employing rigorous quantitative analysis, this study investigates diverse perceptual facets with a sample of 124 participants. Statistical methods, including t-tests, correlation analysis, and effect size measures, were employed to scrutinize participants’ attitudes and behaviors towards AI- generated images in response to varying perceptual factors with the help of Likert scale and AI- generated images. Results: The study unravels critical insights. Trust in AI-based recommendations surprisingly did not significantly affect participants’ attributions of AI-generated images, revealing nuanced dynamics in trust perception. Additionally, heightened ethical concerns notably increased the likelihood of attributing AI-generated images to human creators. Moreover, image complexity exhibited a substantial negative correlation with AI attribution, indicating a cognitive interplay influencing perceptions. Discussion: These findings underscore the complex relationship between trust, ethical considerations, industry beliefs, and image intricacy in shaping attributions of AI-generated content. The implications of this research resonate in the AI revolution era, emphasizing the necessity for a deeper understanding of AI’s potential and its alignment with human attributions. To harness the transformative power of AI, comprehending these dynamics is crucial, ensuring a harmonious and optimized integration of AI in various domains.