抄録
This paper examines the imperfections of Artificial Intelligence (AI)-generated results and the otherness of AI in the authorship of art through discussions of the works of three authors from different cultural backgrounds, who all work with both generative AI and traditional mediums. Throughout history, many artists have tapped into our deepest anxieties, stirring a sense of otherworldliness. Their images often evoke profound and visceral responses, and they linger in our minds long after seeing them. Artists have attempted to create such images in various ways, including intentionally subjecting themselves to extreme physical or mental states or using chance and randomness to guide their creative process. Art that embraces the uneasy and alien feeling not only reflects our inner world but also actively invites us to search for meanings in the imagery. The recent emergence of generative AI has made image creation accessible to an unprecedented degree. However, the generated results, especially with models from the early stages of technology's development, often have peculiar outputs, characterized by a strange, jarring conceptual disconnect between visual elements and have absence of coherent narratives. These peculiar AI-generated images frequently presented viewers with seemingly familiar yet subtly alien elements that triggered a sense of uncanniness and created an unsettling valley between the known and the unknown. The imperfections also reveal the fundamental difference between human creativity and generative AI. Furthermore, the rapid advancement of generative AI in recent years has provoked debates concerning the authenticity and ownership of AI-synthesized visual content and the value of art in our society. The artworks discussed in this paper will be presented in an upcoming exhibition in 2026 with the same title as this paper, and we hope to present a tangible representation of the ideas examined in this paper.
https://ror.org/01j7nq853 
