Computational Power and Subjective Quality of AI-generated Outputs: The Case of Aesthetic Judgement and Positive Emotions in AI-generated Art

This research examines the influence of computational power in AI-generated art on human aesthetic perception and emotional response. A group of 100 participants evaluated a series of artwork either produced by humans or created by AI algorithms operating at three distinct levels of computational power. The study results revealed a general preference among participants for AI-generated art over human-made art. Human participants exhibited a preference for AI-generated art that involved greater computational effort, as indicated by stronger likability and more positive emotional reactions, compared to AI-created pieces that utilized lower levels of computational power. However, the change in perception and emotion is modest on average and showed diminishing returns in term of subjective perception improvement with the increase of the used computational power used to produce the AI outputs. The research underlines critical aspects of AI art and the way it is perceived. Furthermore, the study addresses ethical concerns such as the environmental impact of the computational resources required in generative AI, and the implications this holds for the future of the technology.