Bullshit Genres: What to Watch for When Studying the New Actant ChatGPT and Its Siblings

Abstract

Another communication technology has been introduced, ChatGPT, drawing the attention of many pundits, occupying valuable space on every op-ed page, and inspiring a Hollywood writers’ strike and endless small talk, all steaming a bit with the intoxicating fumes of moral panic or outsized utopian enthusiasm. Research on artificial intelligence (AI) has existed for decades, entering many people’s daily lives in dribs and drabs. ChatGPT and its siblings, however, have focused so many people’s attention on the potential changes that AI could bring to work lives, entertainment, and social relationships that it seems worthwhile to take a moment now in 2023 to discuss what light linguistic and media anthropologists can shed on what is to come. I say this as one of a handful of media anthropologists also familiar with linguistic anthropology who happened to study people’s use of Facebook (alongside other media) only a few years after its introduction to the US media ecology (Gershon 2010). For more than a decade, I have been thinking about how media ecologies change with each newly introduced medium. Here, I lay out what I believe ethnographers of AI who engage with large language models (LLMs) might want to pay attention to in the next couple of years.My starting point is that it would be helpful tosuomen antropologi | volume 47, issue 3, 2023explore how people are responding to ChatGPTin terms of genre, that people’s reactions toChatGPT is to treat it at its core as though it isa genre machine—that is, a machine intelligencethat reproduces and tweaks genres in just theright way for human consumption.