The art status of CG-art from a natural-kind of existence

Abstract

The introduction of Generative Adversarial Networks into the art world has revitalized the classic question: can a machine create art? These algorithms require minimal human intervention to function, so their creations are considered CG-art. In this particular kind of art, computers are not a tool for human service, but an autonomous creative agent. On the other hand, recent cognitive studies have shown that people are skeptical of the idea that a machine can create art. This article aims to show that CG-art cannot be arbitrarily excluded from the art domain. For this, Dickie's distinction between natural-kind and cultural-kinds theories of art is used. Then, using Searle's concept of status-function, I propose to consider art as a language-independent phenomenon. Complementing both approaches, I argue that the best art theory to explain the CG-art phenomenon is a naturalkind theory. This implies that society cannot deny the character of art to these algorithmic creations. Thus, the idea that machines can create art is strengthened.