belowtheline.blog

…and of the Artificial Intelligence that’s subjugated both

Ernő Rubik’s modish mien is undiminished by the years since his thirties, when the Hungarian professor of architecture first led millions of 1980s children to master the art of peeling off and reattaching coloured stickers to his ‘Magic Cube’. Toronto-based Spin Master appears to feel that his famous creation has survived the test of time with equal grace, having recently agreed to purchase the rights to this easier alternative to an Airfix model, for $50m.

For all our differences, younger generations appear to share with mine the need to pretend to be able to unscramble this unscrambleable problem. Although, for today’s more audience-focused teenagers, my contemporaries’ act of disappearing into the bedroom for four hours with a scrambled cube and emerging with a solved one (after playing mostly with something other than the Pritt stick) no longer cuts it. YouTube demands that you’re seen solving the thing. And in minutes, rather than hours.

Luckily, YouTube is also pedagogical home to the cube notation and algorithms that lurk behind the solve. Cubing theory might not sound as seductive as chess opening theory, but it enjoys even greater utility. It also shares everything that chess’s notoriously feral Grandmaster, Bobby Fischer, came to despise of his earlier passion.

Opinion’s divided over the psychiatric diagnostic codes through which we ought to interpret Bobby Fischer’s itinerant later years, as critic-in-chief of chess, and of the country for which he won the world championship in 1972. There’s less disagreement, however, over the delirium of the semitic’s antisemitism — and of his take on the game he left behind.

‘The old chess is dead; played out; it’s rotten to the core.’ You get the idea. Fischer was resolute that championships after his were prearranged. Asked in a 2002 radio interview whether he follows chess at all, Fischer replied: ‘I follow the old chess; I follow all the prearranged matches, like the last Kramnik-Kasparov match. At the highest level it’s all prearranged, move by move.’

🤨

Let’s be honest, though: it’s relatively tame stuff by Outbrain’s standards, and equipped as we are now to handle selfie enhancements, the economic merits of Brexit, and world-beating virus tracing systems without taking any of it too seriously, Fischer’s protest is less offensive to reason than it used to be.

‘You have very interesting, beautiful prearranged games being created by very intelligent players, working with computers, working in teams.’ Triggered no doubt by the realisation that he lacked by then the practice, the mental agility, and the motivation that defined his youth, Fischer looked upon chess no longer as an enduring enigma, but more of a settled matter. Perhaps he recognised the advancement by others of techniques he’d mined himself.

With players now supported by teams of analysts, an encyclopaedic record — and olympian powers of recall — of an opponent’s past moves, together with computer modelling of their next, formula had displaced creative inquiry. Theory and memorisation had pacified innate aptitude. This was an age, he lamented, in which a certain order had been undermined: the coruscating genius of Capablanca dimmed to candlelight by cocksure teenagers with silicon coaches.

Whether or not Fischer’s assessment holds apropos of chess, it’s minimally right when applied to the Rubik’s cube.

The cube is a solved puzzle. It’s played out. Theory and algorithms have rendered this infuriating object, with its 43 quintillion permutations, a memory test. The mythical five-second solve is an impressive feat that requires one to apply a solution, but not to find one. For cubing enthusiasts, there’s aspiration and envy enough regarding the speed with which a known solution can be deployed. For the rest, it’s David Copperfield showmanship masquerading as Newtonian genius; the highly skilled disguised as the divinely inspired. Algos replaced the Pritt stick, and nothing replaced the other distraction.

But isn’t it right for self-improvement to rank over genetic endowment? Industriousness must be of greater virtue than the inequities of natural advantage. If Fischer’s grievance were anything more than a chauvinist preference for nature over nurture, it would have to be premised in the belief that there exists some property of mind so pure that even learning were a contaminant.

The failures of Artificial Intelligence garner fewer headlines than its successes, but in exploring this question they are more instructive. There’s little doubt about the paradigm’s genesis of increasingly sophisticated solutions to increasingly complex problems (having vanquished humanity over both chess and the Rubik’s cube). But A.I.’s proponents’ stubborn abstraction of intelligence as merely pattern-matching and statistical heuristics has left it in permanent exile from the brand of intelligence that created it; those occult creative human faculties, that is, responsible for the startling reality that most of what we say has never been uttered before.

Learned solutions to the seemingly impossible, by both humans and machines, speak only of the ‘how’, and nothing of the ‘why’. They give way to a beauty that deserves to captivate us; but it’s one only of something analogous to John Searle’s ‘Chinese room’, in which ‘understanding’ is not a necessary criterion to formulating an appropriate response. There is beauty in the Machine. But it should keep its place behind that of the Ghost.