Generative AIs, more-than-human authorship, and Husserl’s phenomenological ‘horizons’

Abstract

In recent years, developments in Artificial Intelligence have produced Large Language Models (LLMs) leading to a form of generative AIs (GAIs) trained on vast corpora of texts, capable of producing convincing predictive synthetic texts such as essays, reports, or other texts instantaneously. This development has profound implications for human entanglements with technologies in terms of how these AIs might constitute new forms of subjectivities, texts, and knowledge practices. This development has particular significance for higher education in terms of academic writing, assessment, and research. In terms of higher education, specific concerns have been raised about the implications for assessment, study practices, and the status of knowledge and learning in the context of these ‘writing machines’. Universities and government bodies have reacted in various ways, with some commentators calling for the sector to ‘embrace’ these generative AIs (GAIs) as merely the latest ‘tools’ available for study and research, while others seek to outlaw their use. The current academic research surrounding this phenomenon is underdeveloped with relatively few studies having been conducted, due to the rapid recent acceleration of the technology. It is also under-theorised; early research has reacted to the fast-changing landscape of GAIs by focusing on the technical and practical capacities. Meanwhile, supranational private providers such as OpenAI are influencing higher education internationally, with a complex range of effects which are as-yet unknown. This paper reviews the current state of the art in related bodies of research literature and proposes that the field could benefit from a wider variety of critical and theoretical perspectives. Drawing on the concept of the sociotechnical imaginary from science and technology studies, it considers how discourses and practices surrounding GAIs are evolving in society and education. It then considers the effects of authorship and the writing subject, with reference to the concept of more-than-human authorship. It then draws on recent work in the philosophy of technology, proposing Husserl’s outer and inner ‘horizons’ as a potential framework with which to consider the complex entanglements of human and nonhuman agency, as enrolled in more-than-human authorship and entangled with the presence of GAIs in the ‘lifeworld’ of contemporary higher education. It concludes by proposing future directions for work in this area, in order to gain better theoretical purchase on the phenomenon at the various levels set out above.