Den samtida receptionen och värderingen av AI-genererad litteratur i sverige

Abstract

The contemporary reception of and value negotiations surrounding AI-generated literature in Sweden

In this article, we analyze the reception of the AI-generated poetry collection Ammaseus horisont: AI tolkar Karin Boye (2020) and discuss the more general debate about AI-generated literature in the Swedish press since November 2022, when OpenAI made ChatGPT available to the public. Specifically, we analyze the ways in which Swedish critics construct the literary value of Ammaseus horisont and unspecified AI-generated literature. Our analyses show that different aspects of a text’s origin – especially whether it was authored by a human being, and whether it was published by a publisher with high literary status – significantly affect the valuation process. To capture the ways in which valuation is related to origin, we use the term the origin component of literary value. While critics differ in the extent to which they correlate the origin component of literary value with textually focused values (e.g. style and form, as outlined by two of the authors of this study, Torbjörn Forslid and Anders Ohlsson, in Forslid et al., Höstens böcker: Litterära värdeförhandlingar 2013 [2015]), most agree that the origin and value of a literary text are independent of each other. For example, some critics argue that the fact that a text is AI-generated means that its literary value is lessened, even though it may be beautiful stylistically, while some argue that the same fact can increase the literary value of a text (as a formal experiment), even though the poems themselves are not seen as beautiful. In a larger literary studies perspective, the valuation of AI-generated literature evokes questions of the importance of the author and authorial intent. We discuss such questions in the context of how critics debate them.