Affective computing, algorithmic affect management, and the quantified worker

This chapter investigates what happens when affective work is tracked and monitored, by investigating a subdiscipline of AI research called affective computing, where datafication is in a stage of subsumption previously unseen. Affective computing emerges from computer science departments and practically speaking has to do with researchers’ attempts to both programme computers to recognise human emotions, and to programme computers to display emotions. Affective computing has hitherto not been investigated through the lens of work and worker tracking except by Moore et al. (2024) and by Frank Pasquale (2024). In the current chapter, we set out to adapt the affective computing methodology of coding human emotions by looking at how what we call ‘algorithmic affect management’ (AAM) is likely to emerge where workers’ emotions and affective states are also measured. The newest types of data-driven biometric and other physiological tracking systems are becoming increasingly integrated into working environments to track workers’ emotions and sentiments, as well as physical expressions such as fatigue (which are, potentially problematically, sometimes linked to emotions, but not always).