Ethnographies of Human‐AI Collaborations: What Arrangements of Expertise Emerge?

In this review, I outline four arrangements on the spectrum of human experts‐AI collaboration, as studied from an ethnographic, grounded perspective by sociology, management, and STS scholars. Starting with the sociology of expertise's definitions of what constitutes “expert work,” I first review cases in which AI almost replaced human expertise, those in which AI aims for ‘no human in the loop,’ offering reasons why such complete automation has not succeeded. Second, I review cases where AI is a bit ‘too much’ and becomes exploitive and bossy, where most sociological research is focused, that is, algorithmic management, platform and gig economies, surveillance capitalism, and digital labor. Third, I review cases where AI offers new opportunities for expertise to develop new markets, roles, and co‐productions, and while incurring some professional risks for myself, I title these cases as ‘yay.’ And last, I review cases where ethnographic, grounded studies have shown that experts simply ignore or dismiss AI, with a simple ‘ nah .’ While differentiating these four points on the continuum of human‐AI collaboration, I discuss how the study of AI is entangled with the study of expertise in the 21st century, highlight questions and agendas, and offer directions for future research.