The paper presents research into how academia is addressing the growing issue of valid authorship, given that generative artificial intelligence (AI) tools are now producing similar products to those of humans. In response to this question, the authors undertake an analysis of the differences between what is produced by human authors versus machines(AI) through the examination of these products, specifically an academic essay (i.e., dissertations) created by humans and another created solely by a machine using comparative discourse analytical means. Both products are assessed in terms of the language used to convey meanings and the distinctiveness of the language compared with the author's use of language in their personal writings. The data for this project were collected from thirty-one Papers (i.e., one Paper created using machine-generated text and 30 essays by human-created texts) through an investigation of the organization of the writing, the meaning derived from the use of language, stylistic usage and choice, the way people tell their story, as well as the perception of writers as ideologically positioned. The findings support the conclusion that while the written work produced by machines is functionally coherent, there is less emotional content or risk taken in the telling of these stories. Additionally, there is no real variation among the types of stylistic choices or emotional content within these essays, as opposed to those written by humans, which allude to more grounded and personalized authorship, all of which contribute to a reader's perception of authenticity. Suggesting that there are certain dimensions of authorial voice that cannot be mimicked by AI tools and are therefore present in human-created texts that AI-generated texts do not possess in terms of originality.
Maqbool et al. (Wed,) studied this question.