EVALUATING ARTISTIC AUTHENTICITY IN MACHINE-AIDED SCULPTURES

Abstract

The question of artistic originality with machine-aided sculptures involves a strict, multidimensional study involving the intersection of human creativity and algorithmic determination of aesthetic forms that lead to hybrids. With AI-assisted design, generative modelling and robotic fabrication gaining more and more influence over the practice of sculpture, the issue of authorship, intentionality, material fidelity and interpretive meaning is raised. This paper advances a conceptual and methodological model that combines qualitative expert judging and quantitative computational evaluation to measure authenticity in three categories of sculptures, fully human-created, machine-aided, and fully computer-generated sculptures. The study uses refined collection of sculptures and their creation logs, metadata of processes, and parameters of generative models that allow structural comparisons of human-machine agency. Subtle aspects of authenticity, like narrative coherence, expression intentionality and perceived material integrity are evaluated by expert judgment of artists and curators and critics, and structural complexity, stylistic deviation, and form-generation transparency are computed. Findings have shown that machine-assisted sculptures usually exist in a hybrid zone of authenticity, in which the intent of human action is still present, but an introduction of new aesthetic marks happens through the use of algorithmic forms. Analysis of agreement shows that there is moderate correspondence between expert and computational scores but there are patterns of cultural bias and differing tolerance to the involvement of the machines in audience based interpretation. In general, the paper highlights that originality in artificial intelligence-generated sculpture is not lost but rather reconfigured in new structures that should consider co-authorship, repetitive digital art, and new aesthetic reasoning.