The Nuts and Bolts of Droids and Drones, Bots and Borgs
Since the recent announcement of the opening of a research centre at Cambridge University in the UK for the study of the potential threat posed to human beings by Artificial Intelligence, I’ve been musing on the meaning of “artificial intelligence” — less the hows of it than the whys and wherefores.
The Centre for the Study of Existential Risk (CSER) has also been dubbed “The Terminator Centre”, although the potential existential threat to the human species posed by artificial intelligence is only one of its mandates. Others include the risks posed by climate change, nuclear war, and biotechnology.
Oddly enough, these four existential risks invite comparison to the Four Riders of the Apocalypse in their number, and by extension perhaps even to the activities of William Blake’s four Zoas. What these threats share in common is that they are all potentially perverse outcomes of human intent and formative mental activity itself — the shaping power or faculty. In that sense, the truer purpose of the centre is the study of all those things which I earlier named, following the insights of Jean Gebser, as the signs of the mental-rational consciousness structure functioning in deficient mode — perverse outcome, unintended consequence, blowback, or revenge effect — and which I identified as collectively belonging to the problem of “ironic reversal”. So, this fourfold pattern of perceived risk or threat is quite meaningful in itself.
That we are here dealing with a singular pattern or dynamic of “ironic reversal” is attested to by the language used to describe the purpose of the centre. The anticipation of a ‘Pandora’s Box’ moment invokes the role of Epimetheus. And, you may recall, I earlier named Epimetheus (whose name means hindsight or afterthought), the contrary brother of Prometheus (whose name means foresight or forethought), as the true spirit of Late Modernity or post-modernity. In addition, there is a very close correspondence between the symbolism of Epimetheus and Pandora with that of Adam and Eve. The trajectory from Promethean Man to Epimethean Man describes the arc of ironic reversal, just as the passage from Parsifal to Don Quixote describes the arc of ironic reversal marking the beginning and ending, rise and fall, of the High Middle Ages. In Parsifal, we have the fool who becomes a knight of the Grail. In Don Quixote, we have the knight who becomes a fool once more, the anachronistic aristocrat who is mercilessly ridiculed and mocked by the ascendant bourgeoisie. Le ridicule tue.
It is easy to identify the beginnings of our own process of ironic reversal and the onset of the Epimethean age, as it were. It is the First World War and the disillusionment of the Age of Reason and the premisses of the Enlightenment. It is, after all, Prometheus who steals the fire of consciousness from the gods and bestows it upon man, setting the world alight. The Enlightenment was this fire in the mind, and was the self-consciousness of Promethean Man. But that light appeared to many to have been snuffed out by the War and its consequences. After the world wars, no one could take seriously the optimism of a Condorcet who earlier announced the Enlightenment faith in “the infinite perfectibility of man” through the judicious application of universal reason.
Which brings us to the why of AI. It is not entirely a new concept, as it has its roots in old legend and ancient magic, as the golem, for example, and of the magical animation of the inanimate. Goethe’s tale of The Sorcerer’s Apprentice is often invoked as a cautionary tale about the perverse outcomes and unintended consequences of wayward magic or magical malpractice. There is, in this sense, a very close association between what is called “techno-science” and what Gebser calls the “magical structure of consciousness” whose common denominator is will to power, and the desire to command powers and forces to do one’s bidding. The problem arises when one is not fully conscious of either one’s own true motives and intents or of the full nature of the forces and powers one seeks to command. Interestingly, the latter-day sorcerers who do achieve full mastery and command of such powers and forces, such as Castaneda’s don Juan or Gurdjieff, have virtually no interest in exercising it, dismissing it as mere “tricks”. Power, for don Juan, was after all one of the “four enemies of the man of knowledge“, along with fear, clarity, and old age, that had to be overcome in order to arrive at the true goal — total freedom realised in the awareness of the “totality of oneself”. Power was a diversion and distraction from that goal.
So, there is a large dose of mysticism and old magic underlying the motivations for pursuing artificial intelligence of which those who pursue such goals are virtually oblivious, yet which, ironically, animates them and which subliminally guides their thinking and their activity. That becomes evident when you watch promotional videos such as “2045: A New Era for Humanity”. It may be “new” but that doesn’t necessarily make it “good”. Means and ends remain confused. The pursuit of newness in itself — innovation as an end in itself — has become a kind of mental automatism or trance-like march of the zombies. The constant pursuit of the new without having even completely penetrated, digested, or understood the “old” is a formula for trouble. It is, in fact, what got the sorcerer’s apprentice in trouble.
But, of course, techno-science dismisses such comparisons between scientific innovation and the sorcerer’s apprentice. We are assured that there are built-in safe-guards against such activity running amok. But there is an essential self-negating contradiction in that assurance in relation to Artificial Intelligence. The incentive for pursuing technological and practical applications of artificial intelligence is the need for new devices that can coordinate and operate in an information and social environment that has become mentally unmanageable. As we discussed earlier, one aspect of “deficient rationality” as it has emerged is that the mind or intellect is no longer capable of mastering its circumstances, circumstances it has largely generated from its own activity. The linear, syllogistic, cause-effect oriented intellect is ill-suited to “everything all the time” and the all-at-onceness of the global era. Artificial superminds are seen, therefore, as a solution to the problem of coordinating and managing everything-all-the-time. So the very thing that techno-science assures us is impossible — loss of command and control — is actually the very thing which exists in fact as the motivation for designing artificial superbrains in the first place.
And that self-negating contradiction is what actually came to the fore in the recent market meltdown, in which AI-type algorithms used in the finance sector, performing millions of transactions per minute, ran amok. And as these devices progress in complexity, we can be quite certain they will escape and overrule similar built-in safeguards, because the very assumptions underlying their development, or which provide the rationale for the development of artificial general intelligence overall, negate the self-conscious, ostensible intentions or motives of their designers. And it’s hard to understand how that contradiction can be overlooked unless it is deliberate deception.
In the next post on this subject of the nuts and bolts of bots and borgs, I’ll want to address in more detail the symbolism and psychology behind artificial intelligence, as well as the abuse of the meaning of “intelligence” that is implied in the phrase.