Artificial intelligence and the ethnographic encounter: Transhuman language ontologies, or what it means “to write like a human, think like a machine”

Abstract

In this paper, we employ the language ontologies framework to artificial intelligence (specifically, OpenAI's ChatGPT) to investigate the ‘ethnographic encounter’ between human and non-human language users. Our focus is on the exchange and interplay between human language users and non-human artificial language generators in the production of written text. We analyse how such programs transform our understanding of what language is or might be; their practices to create language are unfamiliar, and yet they make sense to human interlocutors. Drawing from, and building on, the language ontologies framework, we discuss the practices involved in such encounters and suggest the need for an updated ‘toolkit’ in our understanding of language to account for transhuman interactions.