Enchanted determinism is when a technological system is described in terms of the sublime. It is the myriad of issues that Campolo and Crawford note that problematize what is at stake, which is that enchanted determinism does not give way easily to a disenchanted scientific rationality. They write that in the enchanted world, where data is unreasonably effective at defying our disenchanted intuition, predictive efficacy trumps casual explanation, and theorization or casual explanation can be dismissed. Mythmaking comes in to save, sell, the process. Its enchantment a shield for its creators to hide from accountability, whether legislative or social. And determinism is its weapon, as when these deep learning systems are deployed in critical social areas, new problems arise out of what is not known about how systems enact forms of control and territorialization.
The true insidious nature of enchantment doesn’t work if AI is snake oil, it’s why Campolo and Crawford recognize that enchanted determinism operates when these systems succeed at providing a hyped process where magical mastery and technical mastery fuse. It goes beyond marketing’s role in producing myth, enchantment is intrinsic to the advancement.
A core point of their article is this linkage of AI and alchemy as a field of research. This conceptual pairing left scientific respectability at the door, even if corporations still always knocked. AI work as alchemy is just the current example of the engineer, which is not so much a historically stable discipline, but one in contrast to the modern disciplines. In Zeroes and Ones by Sadie Plant, she writes about Engineering as a field that travels on experimental routes, which throwback to the skills of lost shamanic cultures, the trials and errors of alchemy, and brews condemned for witchcraft in the centuries before the Enlightenment.
While the inception of engines in the mechanical age might make it clear its inception point as a discipline, engineering is not confined to the production and use of machinery in factories (this industrial birthing point, of which, is an indictment of the technological determinism embedded in the field, in a very Oedipal way). Plant writes that engineering may have been defined here among the levers and automata of the 18th century, but the genealogy was not invented here, parentage points further backwards.
According to Plant, what most interested Freud about Leonardo Da Vinci (who is often said to be the West’s first engineer, Archimedes?), was not Da Vinci’s ability to capture the essence of femininity in his art, but his “alien interest in experimentation”, which brought him “close to the despised alchemists, in whose laboratories experimental research had found some refuge in those unfavourable times. There’s always been a—albeit ostracized—home for engineering, but an economic system for it? That’s moderately new. While Plant wrote this in 1997, her indictment of the engineer rings true in the context of the deep learning “scientist”:
“Engineers are not the authors of anything, but simply technicians and caretakers, carrying out instructions written elsewhere and looking after the machines entrusted to their care.”
Technical invention is a field of bricolage, a tinkering improvised human endeavour based on diverse miscellaneous pieces, which operates outside of scientific argumentation, and not linked to social demand.
Crawford and Compolo strongly lay out the postmodern criticisms of AI, which designated it as a protoscientific and premodern science, that which was intended to detract from its respectability as a science now seem to have only made it stronger. When faced with black boxes of complexity, enchanted determinism is the onus we give on processes we can’t fully comprehend. Because, it almost works better if we attribute the mystical to it, offhanding your social agency feels better if you project it as giving away to a higher power. AI is an irrational rationalization that we attribute as sublime superhuman intelligence, when it is just complex statistical modelling and prediction sustained by inconceivable stacks of information (which are contextless). It intensifies hierarchies, differences, marginalization, and closes off the accountability that could meet it.
And one wonders why the world flies by at a frustrating rate, society is marked by the cultural change of the impossibility to get a grip, to feel the changes, before they are changed again, advanced again. Hyperstition, etc. This ability to know and control changes manifested on the social scale was crucial to modern conceptions of understanding humanity’s place in the grand scheme of things. Technology could be seen as an attempt to realize this power, but that quickly slipped out of hand. Now we are enchanted with the tools of our displacement, and techno-optimism is a resignation of our place in time.
References
Campolo, A., & Crawford, K. (2020). Enchanted Determinism: Power without Responsibility in Artificial Intelligence. Engaging Science, Technology, and Society, 6, 1–19. https://doi.org/10.17351/ests2020.277
Plant, S. (1997). Zeroes + ones: Digital women + the new technoculture (1st ed). Doubleday.
-> Slopunk <- 