Key points are not available for this paper at this time.
AI has captured the artworld, and, increasingly, humans' engagement with many forms of media. Computer-generated art sells for millions at auctions; artists routinely use algorithms to generate aesthetic materials. However, to capture the impact of such works and our relationships with them, we need to better understand the kinds of responses we make to AI/computer-generated images. Here, we consider whether and, if so, to what extent humans report feeling emotions when engaging computer-generated art, or even ascribe intentionality behind those feelings. These are emerging—and also long-standing—points of controversy, with critical arguments that this should not occur, thus marking potential distinctions between artificial and ‘real’ human productions. We tested this by employing visually similar abstract, black-and-white artworks, made by a computer (RNG) or by human artists intentionally aiming at transmitting emotions. In a 2 × 2 design, participants (N = 48) viewed the art, preceded by primes about human/computer provenance (true, 50% of cases). Contrary to critical suggestions, participants almost always reported emotions and ascribed intentionality, independent of the prime given. Interestingly, they did report stronger emotions when the work actually was made by a human. We discuss implications for our understanding of art engagements and future developments regarding computer-generated digital interactions.
Demmer et al. (Fri,) studied this question.