On AI, Contemporary Capitalism, and the Ascesis of History
I apologize, dear reader, for neglecting this newsletter for so long; I’m aiming to get to a weekly or even twice-weekly schedule and will aim for a revival of the Friday newsletter, along with more frequent occasional thoughts and partial essays, if you’ll forgive me the luxury of airing my half-finished thoughts here. Earlier this month I decided to step back from my social media presence and am trying in general to reduce my casual consumption of news and commentary; so far it has had a pretty salubrious effect. Speaking of social media, when I started out writing here one of my goals was to discuss technology, including social technology, and its political, spiritual, and historical implications and aspects; I have not thus far come through on that intention but hope to begin making up for it. Look for further explorations of technology present and past in the months to come.
My professional work of late has intersected with the vast world of machine learning, AI, data generation, and other associated practices and bodies of production and knowledge; my very small contribution has been towards development and refining of optical character recognition technology for Arabic script, to be followed and already in the initial stages in fact, of similar work for handwritten text recognition in Arabic script manuscripts. Besides the technical aspects of this work I have reflected at length on the philosophical, ethical, epistemic, and other implications and inflections of this kind of intersect of digital technology and methodology with ‘traditional’ humanities concerns and resources. Such reflection is, I think, by necessity two-fold: there is the specific work being done and the possible implications within the rather rarefied world of Islamicate academic studies and life, work and immediate ramifications over which we—myself and my colleagues, not an especially vast number!—have a considerable degree of control and say, albeit far from absolute. Second though there is the wider, indeed global, context, in this case of the technologies on a large scale and the social, economic, and other structures involved. One of the most looming aspects of this wider context is the question of artificial intelligence, its capacities, possible future capacities, and the ends to which it is being put and which, perhaps more importantly, powerful people and institutions imagine it will be put towards in the future. What follows are my attempts to begin making sense of these wider implicaitons of AI as it is imagined by many and as it and other emergent digital technologies are being deployed.
There appears to be a very real sense in which the drive for AI and for various forms of total or near total automation of a highly sophisticated nature is not meant to generate ‘innovation’ per se but rather the latest—final, last-ditch?—effort to make up for the general exhaustion of resources that is so typical of industrial capitalism and its insatiable hunger for more of everything, a hunger that periodically threatens to outstrip the productive outputs of the global machine. The promise—chimerical or otherwise—of AI is for many that it can replace human intelligence on many fronts, and this is necessary, implicitly if not explicitly in these calculations, precisely because the efforts and depredations of state and capital have reduced human intelligence, contained it, neutralized it, and generally starved the kind of full-front development that previous modalities of state and capital both developed and grew out of. Simultaneously those same massive expansionary energies of the post-War economic, political, climatic, etc order have generated problems, including not least of all just immense data sets (real and more metaphorical), at such scales as to preclude even the most ideally developed ‘human resource’ base. In a more reasonable and sustainable world there would be no need for AI beyond fairly limited applications, as extensions and supplements to human intelligence, exploring questions of scientific interest that involved scales either exceeding human capacity or operations of a most tedious and dull nature such that their replacement would—one hopes—not involve any serious trade-offs intellectually. The real dream, however—and it may never actually move beyond dream—is of AI that will outright replace human minds (and the bodily operations of those minds, in this imagining), that will make all manner of human capacity and labor and intervention redundant, or, as seems increasingly clear, would substitute for human capacity that has been either eliminated or otherwise overwhelmed by the outworkings of industrial capitalism and state systems.
On almost every front capitalist innovation is in a scramble to deal with the accumulating effects of capitalism itself; of course to a considerable extent this is not new, but has been the story of industrial capitalism for a long time, as one resource base after another is exhausted, or as new social and political problems are generated by the operations and very successes of capitalism and that must then be solved, but it has I think become more fundamental in recent years, heightened by the fact that there are few problems (in the socio-political etc senses, at least) left truly external to the industrial capitalist system: everything has been to one degree or another consumed and incorporated, and what we call the ‘anthropocene’ is really the world of industrial capitalism and state formations, in a particular trajectory at that, not simply ‘modernity’ writ large and total. To be sure the rate and degree of consumption and absorption is not the same everywhere, either geographically or in a non-spatial sense, but at most this ‘failure’ of total consumption means that there is still friction and possibility, albeit highly reduced and tenuous.
Capitalism itself is incapable of truly addressing any of the crises it has generated, and the same is true of the modern state system; they are the crisis, fundamentally, and so long as we persist in expecting solutions purely from within them we will be disappointed. This is not to say that everything that we identify as ‘modernity’ must be discarded, if that were even possible—and it is not. Rather, what is truly necessary is a critical genealogy of modernity and attention to the multiple routes and inheritances that have passed into the present, to trace backwards to alternative modernities, to different branching points, and so to realize what are the actual alternatives and possibilities going forward, what aspects of modern systems we are stuck with, which have promise for humane and constructive future development, and which are inextricably bound up with the current system of industrialized state and capital and ought to be resisted and ultimately discarded or replaced. We cannot get outside of modernity or even of industrialization or of the systems which we inhabit and which to a great degree inhabit us, but we can begin by imaginatively exploring both the parameters of our own world and of previous worlds, catching glimpses of what could have been and what might still be in the process.
History, then, as a discipline—and I am thinking of both human history as we tend to think of the discipline now conventionally, but also of natural history as an integral part of any true and good project of history—has the vital role of taking us out of ourselves, of reducing our focus on our self and what is natural and normal to the self as it exists now, and putting us in the presence of others and of other times and selves and beings and ways of life. History becomes a type of dying of the self and a relativizing and distancing of the present and of one’s little internal world as generated by one’s self and one’s immediate contemporaries and the socio-economic etc structures that help to constitute and which surround the self. In this sense history is almost an ascetic discipline, an act of self-negation in refusing to collapse the past and the being of others, human and non-human, into one’s self and one’s own particular concerns. Examination of the past opens one up to other possibilities, possibilities which might unsettle the concerns and desires of one’s self and make one realize just how much what we think of as our autonomous and self-generated reality to be the work really of contemporary structures. There is more that history—human and natural—can do from such a point, though it reaches its limits, as it cannot of itself generate the act of return: that is, the mystical going into one’s self for transformation and communion with God and with others.
The discipline—and I have in mind here very much the old sense of ascetic action, of history as an ascesis—of history is primarily preparatory and initially generative, powerful to be sure and I think particularly in our world necessary, but far from sufficient or final. To loop back around to our initial topic, then, we can, and probably must, incorporate a wide range of tools in the pursuit of this discipline (and of others, for that matter, though this is not an absolute prescription), including digital technologies such as even AI—but in a manner that is fully aware of the wider implications and contexts of those tools and that does not seem them as harmless and value-neutral but as requiring the active incorporation of value and constant guidance and a type of conscious resistance. Otherwise the tool will tend to assume a sort of agency and will collapse one’s work and life into the larger ‘machine,’ no matter what one’s good intentions might have been setting out.