Causal and Communal Factors in a Comprehensive Test of Intelligence
The paper traces a pathway through the existing space of argumentation surrounding the original Turing Test (TT) and the discipline of ‘Strong’ Artificial Intelligence that followed on from Turing’s work, and extends this path to motivate a strengthened conclusion regarding what would be required by a truly Comprehensive Intelligence Test (CIT). The paper begins by examining the initial ‘intelligence test’ as proposed by Turing, and Searle’s high profile critique of both the TT and Strong AI. In tracing the ensuing dialectic between Searle and his own critics, I support Searle’s rejection of the ‘Systems Reply’, and for reasons based more on the philosophical views of Putnam agree that the original Turing Test is fundamentally inadequate. The situation becomes more complex with the ‘Robot Reply’ and allied Total Turing Test (TTT), and I argue that Searle’s attempted refutation of the combined Robotic-Systems reply is unconvincing. However, this is not to say that the position expressed by Searle’s opponents is itself confirmed, and I argue that externalist views in the philosophy of language first put forward by Kripke and Putnam cast serious doubt on the issue. In turn, the causal and communal factors highlighted by externalist views in the philosophy of language point to the need for a fundamental shift in conceptual perspective and a strengthening of criteria in a truly Comprehensive Intelligence Test. I argue that an ideal CIT should focus on the category of cognitive system as a whole, rather than on the performance of individual artefacts. From this expanded perspective, the central question is not whether an isolated agent could simulate human performance within the context of a preexisting sociolinguistic culture developed by the human cognitive type. Instead the key issue is whether the artificial cognitive type itself is capable of producing a comparable sociolinguistic medium of intelligence, where this essential medium is simply taken for granted as a precondition of the individual performances evaluated in the TT and TTT.