A.I. go by many names: towards a sociotechnical definition of artificial intelligence

Abstract

Defining artificial intelligence (AI) is a persistent challenge, often muddied by technical ambiguity and varying interpretations. Commonly used definitions heavily emphasize technical properties of AI but neglect the human purpose of it. This essay makes a case for a sociotechnical definition of AI, which is essential for researchers who require clarity in their work. It explores two primary approaches to define AI: the rationalistic, which focuses on AI as systems that think and act rationally, and the humanistic, which frames AI in terms of its ability to emulate human intelligence. By reconciling these approaches and contrasting them with landmark definitions, the essay proposes a sociotechnical definition that includes the three central aspects of i) technical functions, ii) human purpose, and iii) dynamic expectations.