First define 'intelligence'. OED: 1) faculty of understanding, 2) quickness, 5) knowledge; exchange of knowledge. To abstract from Chalmers: cognitive software

First define 'intelligence'. OED: 1) faculty of understanding, 2) quickness, 5) knowledge; exchange of knowledge. To abstract from Chalmers: cognitive software supervening on a conformal neural hardware substrate. OK but in everyday empirical-focused psychology, how are subjects tested for 'intelligence'? My brother and his colleagues draw from Gardner's eight 'abilities': rhythmic (musical), visual (spatial), linguistic, logical (mathematical), bodily (kinesthetic), interpersonal, intrapersonal, and naturalistic. IE, there are subjective elements in evaluating 'intelligence' plus 'fuzzy logic' only supplies a useful fiction for quantifying the subjective. Now this article appears preoccupied with mapping-out neural hardware specs in its emphasis on processing capacity, memory, clock speed and whatnot. Would a savant mentally capable of calculating pi to a zillion places but incapable of communicating results be judged 'intelligent'? MRI maps of mental processing only would suggest such. Would a quantum MB capable of running MATLAB at warp speed but unable to render output in an ascertainable form be judged well-scripted? Not! Even a perusal of the current WAIS-R manual (no particulars since I'm not predisposed to be shot) exhibits classes of both objective testing and tasks which would be categorized also as measures of interpersonal engagement (Who was famous for ...). Would the Ava character in the movie 'Ex Machina' be judged consciously 'intelligent'? Consensus probably would be 'yes' since both her subjective and objective 3D GUI seemed correct by human standards towards the end. [wonder if anyone who saw the movie concluded that the affective agent of Ava’s cognitive development harkened back to John Locke's premise that percepts functioned as something analogous to self-extracting .zip files which imposed real cognitive patterns on a subject's 'tabula rosa', not requiring any native cognitive modeling]. But would Ava really 'understand' anything or just be a 'Searle's Chinese box'?