“…And also from a system engineer's perspective, given that at any point in time there will only be a finite number of commonly accepted tests of intelligence available, the maxim underlying nPAI seems dubious: Each of the individual capacities could be addressed by a specifically dedicated standalone subsystem, so that intelligence would actually be reduced to correctly selecting and executing the respective module from a finite number of available subprograms based on a finite number of possible input categories. But this could hardly be considered satisfactory as an answer to the intelligence puzzle for anyone but diehard followers of Descartes: At least the Cartesian would be comforted in that the resulting AI -being an almost ideal instantiation of a "type-c machine" [5] -might be able to pass any test for a particular set of cognitive capacities, whilst still failing a test for any mental power whatsoever. General Psychometric AI: A second, more commonly used definition of PAI, to which I will henceforth refer as "general PAI" (gPAI), is also introduced in [1]: "Psychometric AI is the field devoted to building information-processing entities capable of at least solid performance on all established, validated tests of intelligence and mental ability, a class of tests that includes not just the rather restrictive IQ tests, but also tests of artistic and literary creativity, mechanical ability, and so on.…”