On Computing Machinery and Intelligence

Abstract

The Turing Test has been part of the lexicon of artificial intelligence ever since Turing proposed it in his famous paper, “Computing Machinery and Intelligence” (1950). Close reading suggests, however, that Turing’s real purpose in writing the paper was, firstly, to attack the arguments of skeptics so as to establish that there is no reason to believe computers cannot be intelligent, and secondly, to propose a program of research. Today, Turing would likely have structured his thinking differently, and perhaps focused on different questions, perhaps the questions on which I focus: What is it that makes us different from all other species? And what is it that we have in common with other species that makes the difference important? I conclude that story understanding makes us different and that story understanding rests on directed perception. I elaborate on story understanding, explaining how a simple substrate of English analysis, common sense inference, and concept search enable the Genesis story understanding system to demonstrate a range of competences, including culturally grounded story interpretation and question-driven analysis. All this leads to a discussion of open questions and a reassessment of Turing’s paper’s fundamental contribution.