Will your boss be an algorithm? A patent-based analysis of artificial intelligence worker management technologies and labour exposure
摘要
This paper assesses and quantifies the extant development of AI worker-management technologies, intended as AI-enhanced algorithmic technologies that execute functions typically in the realm of managerial decisions. We provide novel empirical evidence on their degree of penetration via an identification strategy based on patent technological classification. By means of natural language processing, we uncover specific scopes of application, and identify human tasks and occupations most susceptible to technological exposure. Results point at a context-dependent exposure, of both managerial tasks subject to substitution, but also of managerial functions enhanced by these technologies. Subordinated activities conducted by mid-low hierarchies tend to be highly targeted in terms of control and monitoring, while they fail in providing amelioration of health and in mitigating safety risks.