Artificial intelligence, particularly after the recent explosive advances and widened uses, has fired up the (previously quiet) debates about the nature of reasoning, creativity, and agency. This paper examines these issues through the lens of classical kalām (Islamic theology), with some focus on Muʿtazilite principles. It begins by presenting a concise overview of the major schools of kalām (Muʿtazilism, Ashʿarism, and Māturīdism), highlighting their respective treatments of reason (ʿaql), divine creation, and human action. Then a brief review of modes of reasoning is provided, shedding light on differences between human and artificial reasoning, stressing the distinction between statistically generated outputs and contextually grounded, meaning-oriented cognition. Then, drawing on Muʿtazilite conceptions of reason, objective morality, and true human agency, in particular, the paper argues that contemporary AI systems, despite their impressive capabilities, do not satisfy the conditions for knowledge (ʿilm), creation (khalq), or agency (fiʿl) in the theological sense. It is argued that although they may appear “creative” or displaying origination (ihdāth) capabilities, AI systems, so far and to the extent that current developments seem to indicate, lack the essential features of ʿaql (reason), nafs (soul), rūḥ (spirit), and niyyah (intention) that Islamic theology identifies as the true, defining aspects of human beings.
Nidhal Guessoum (Thu,) studied this question.