On a Possibility of Artificial Reason: J. McCarthy, I. Kant, and A. Turing

Abstract

The purpose of this study is to explore the possibility of reconciliation between Kant’s transcendental idealism and McCarthy’s epistemological point of view on artificial intelligence, which are at first glance likely to be considered contradictory. For this, characterizing the standpoint of J. McCarthy, who coined the word ‘artificial intelligence’ as scientific realism and that of A. Turing, who provided a crucial thought experiment that shaped the contemporary conception of artificial intelligence as behaviorism, we shall compare these two standpoints with the transcendental idealism of I. Kant, who conferred on us a monumental indicator for understanding the human reason. Through this comparison, we shall argue that scientific realism, which is currently a prominent philosophical standpoint of artificial intelligence, is not compatible with Kant’s transcendental idealism but assumes a standpoint strikingly analogous to behaviorism. Nevertheless, we shall also argue that once transcendental idealism is looked at from the viewpoint of behaviorism, scientific realism can be seen as compatible with transcendental idealism. This compatibility we name the possibility of artificial reason in this paper.