Quantitative definitions; and, would you like to play a game?
12 January 2008
A striking contrast is presented between the new MacPro that just arrived at work and the definition presented for the first learning tic-tac-toe playing computer laid out in a 1963 paper and reiterated in the O’Reilly Statistics Hacks book I’m reading now. On the one hand, an eight-core CPU, new fast RAM, and an embarrassment of riches in graphics cards; on the other, to quote Bruce Frey quoting Donald Michie, 287 matchboxes and large numbers of nine different colors of beads. That’s right. The arch-nemesis from 1983’s WarGames can be constructed from the contents of your kitchen junk drawer.
The argument lies, then, in what constitutes a computer. The MacPro that one of our customers ordered and I got to play with today is by itself, to paraphrase Dave Byrne about telescopes, some rocks and sand. Only with the addition of the complexities of firmware, operating system, software, and, indeed, shimmying electrons, does it become the marvel of basic arithmetic that will edit video faster than Weird Al in UHF.
Concomitantly, the 287 matchboxes and large supply of nine colors of beads only give you a run for your X- and O-money when supplied with a set of rules. To wit, each box is labeled with a particular configuration of moves. Then, those same boxes are supplied with beads, assigned colors to squares, representing legal moves. In other words, the box with an empty board drawn on it will have all nine colors within. A box with two X’s in opposite corners and an O in the center of the board will have six beads, representing the six open squares bifurcated by the line of already chosen moves.
Takes longer to describe the latter than the former here, but I’d bet that not only is all the stuff it takes to make the towering iProduct run more complicated than the ground rules for the automated-naughts-and-crosses-opponent, but that it is mathematically provably so. I still haven’t quite figured out how widely applicable Godel’s incompleteness theorem is, but evidently it extends beyond the rigid logical framework it was conceived in. Might be applicable here.
In the introduction to Amir Aczel’s The Riddle of the Compass, he describes said gadget as the first quantitative measuring tool, setting us along the path to ever more precise gauges and meters. I’m not so sure that that is an admirable course to chart, if you can forgive the metaphor. The slogan adopted by one astronomical observatory is ‘We teach humanity about the universe.’ If you brought Caravaggio back from the dead and sobered him up (which might take the longer of the two), I bet he’d say he was attempting something similar. Zhuangzi and bell hooks are also among those who perhaps accomplish more for not having arrogated quite so much in their public statements of objectives. Ultimately, I’m not sure who among all of them is most effective. That’s not to say that I am disparaging astronomy in general, just trying not to let it set itself on a pedestal. That oughta get some comments going.