Оригиналност и стилова неопределеност в изкуството с изкуствен интелект. GAN и CAN в естетическа перспектива.
Originality and stylistic indeterminacy in ai art. GAN and CAN in aesthetic perspective.

Summary/Abstract: This article examines the impact of Generative Adversarial Networks (GANs) and their modification, the Creative Adversarial Network (CAN), on concepts of originality and style in art. Employing an interdisciplinary approach that combines theoretical reflection and empirical analysis, the study evaluates the stylistic and aesthetic differences between images generated by the two technologies. The empirical investigation, based on expert evaluation of 60 images, reveals significantly higher levels of originality and stylistic ambiguity in CAN-generated artworks. The results suggest that CAN not only imitates existing styles but also creates novel visual forms that challenge historical paradigms of stylistic development. The article further addresses the cultural and ethical implications of AI-generated art, including issues of authorship and bias in training datasets. The study contributes to contemporary debates on creativity and machine generated art in the age of algorithmic production.

• Details
• Contents