Take a Chance on Me: Aleatory Poetry, Generative AI, and the External Demarcation Problem

Abstract

Abstract What is it in virtue of which any poetic output will be included or excluded from the category of art? I will first identify the external demarcation problem, which is concerned with whether or how the cut-off is made between art and non-art. I will then adopt a nonclassical approach to conceptual analysis by relying on a set of examples of poetry generated by aleatory processes to evaluate an intention-based response to the external demarcation problem. I will argue in favor of an intention-based response that is grounded in hypothetical intentionalism. According to this response, a contextually informed audience will form a hypothesis about poetic intentions on the basis of the evidence that a work makes publicly available. Semantic, categorial, and ostensive intention and intention traces may help this audience to determine whether a work counts as art and is worth effortful interpretation. My proposed version of an intention-based response to the external demarcation problem will be based on the p-valued hypothesis-testing approach in science and will be highly relevant to a context of production in which we find human poets, poetry-generating AI systems, and human-AI interfaces.