Information systems research on artificial intelligence and work: A commentary on “Robo-Apocalypse cancelled? Reframing the automation and future of work debate”

In his article “Robo-Apocalypse Cancelled? Reframing the Automation and Future of Work Debate,” Willcocks provides a critical reflection on the common assumption that artificial intelligence (AI) technologies have a massive impact on jobs. The article offers a much-needed and very helpful discussion entry that I certainly recommend to include in our teaching. While I fully agree that it is time to cancel the RoboApocalypse, I will argue that we first need to get rid of the more general and persistent naïve anxiety and admiration of the power that we assign to AI technologies, what Willcocks describes as the hype and fear narrative. It seems that we tend to ignore the most urgent question: what is it exactly about these technologies that legitimates all this fuss about AI changing our work? I will argue that it is time for information systems (IS) researchers trained in the sociotechnical tradition to step on board and fight the hype and fear narrative by offering empirically validated insights on why, when and how AI changes our work. Let me start with a disclaimer. Current research on AI technologies and work does not only address job losses. In fact, over the past 1 or 2 years, scholars from different disciplines have been studying changes in the quality of work instead of quantity of jobs. While strikingly silent during the Robo-Apocalypse debate—fearing among other things a second AI winter—computer scientists have recently joined the discussion on AI and work, stressing the need to keep the human in the loop (e.g. Dellermann et al., 2019). By developing hybrid AI, tools will become our new assistants, coaches and colleagues and thus will augment rather than automate work. While computer scientists’ aim is to build systems that increase the quality of work, a recently growing group of critical researchers aim instead to create societal awareness about the rise of low quality of work due to AI. They point, for example, to jobs under constant surveillance (Zuboff, 2019), jobs characterized as “ghost work” (Gray and Suri, 2019) and jobs imposed by inscrutable data-driven decisions (e.g. Faraj et al., 2018). In addition, organizational behavior scholars mainly use survey-based research to analyze, for example, the individual perceptions on the quality of work with AI (e.g. Brougham and Haar, 2017). Even those who started the hype–fear narrative, principally labor economic scholars, conduct research that departs from the focus on changes in number of jobs. For example, Felten et al.’s 2018 study how AI technologies are associated with changes in skills and wages and found that these new technologies might trigger a different type of polarization than one related to job losses and gains. While these latest academic contributions address the changes in quality of work instead of quantity of jobs, they share the implicit technological deterministic assumption that AI has the power in itself to change work. This is rather problematic; while technology has its own agency, it is also always socially constructed, which makes assumptions about its decisive impact unattainable and will only sustain the myths surrounding AI. Based mainly on the tradition of the field of Sociology of Technology Studies and Actor Network Theory, sociotechnical IS researchers know how to go beyond this technological determinism (Cecez-Kecmanovic et al., 2014). By using an historical and multi-actor perspective, Information systems research on artificial intelligence and work: A commentary on “RoboApocalypse cancelled? Reframing the automation and future of work debate”