The artificial author is half-constructed: Historicizing generative AI text

Moral panic in higher education and writing communities prompted by recent developments in generative artificial intelligence (AI) has focused on how large language models (LLMs) facilitate plagiarism, cheating and writing scams. Dialogue around managing the existence of LLMs centres on traditional modes of solo human authorship. Such discussions often overlook the collaborative potential or what Scott Rettberg terms ‘cyborg authorship’ opportunities presented by LLMs. But while programmer/authors of computer-generated texts of decades past were typically self-conscious and transparent about their machine authorship, the advent of AI chatbots presents a new dilemma for the notion of authorship because of their seamless integration into writing practices and lack of transparency around their algorithms. Yet, the history of computer-generated texts and authors of electronic literature offers examples of how conversations around AI chatbots can be reframed to provide greater literacy about what constitutes computer-assisted and computer-generated writing and the creative potential of human–machine collaborative authorship.