Jason Q Han
Jason Q Han is a 2nd year PhD student in the English Department at UW-Madison. His current interest involves depictions of solitude and loneliness in 20th and 21st century literature and philosophy.
The uptake and integration of ChatGPT and similar large language models is so ferocious that one would be mistaken for forgetting that its first broad consumer release came in 2022. On top of mammoth energy and water costs,[1] current financial markets have made what is often described as a “big bet” on AI technology[2] to transform everyday life in a manner similar to that of computers and the telegraph. How AI will accomplish this ambition precisely is unclear, meaning there exists today an enormous speculative asset with little clear sense of its applications. Before the “AI moment,” the logic and ideology tasked with its unfolding guarantee a further etiolation of the terrestrial circumstance. In the former, the profit motive. In the latter, transhumanist libertarianism. An emancipatory AI future requires the disintegration of both. Furthermore, we might find a possibility for an alternative only according to a different discourse. The task for us, or for those not-quite like us, is the same as the one forwarded by Gilbert Simondon, who wrote, “The collectivization of the means of production cannot achieve a reduction of alienation on its own; it can only achieve this reduction if it is the precondition for the acquisition of the intelligence of the individuated technical object by the human individual.”[3] The possibility of Simondon’s “technical culture” requires therefore an encounter with AI, a contemporary technical object at the forefront of hip financial speculation and mystification.
At this point we may introduce Anne Alombert’s Digital Schizophrenia & Other Essays. Its signature theme, drawing heavily from Simondon and Bernard Stiegler, is to shift the AI question. First, in sentiment. For Alombert, part of the AI problem is the cultural “schizophrenia” between capitalo-transhumanists and those critical of AI based on supposed affronts to the sanctity of human life. Instead, it is worthwhile in the present to critique AI, not by virtue of what it is, but rather what it does, its material effects in the world. An encounter with this question however is also obfuscated, according to Alombert, by the particular discourse crystallized around AI. This discourse finds its origin in Neoliberalism and cognitive science, and its main problem is consciousness, that is whether AI can think. Beyond questions of consciousness, which tend to rely on “problematic analogies between mind and program or brain and computer,”[4] Alombert instead articulates a genealogy of the “externalization of minds,”[5] a problem intrinsically related to humanity’s technological embeddedness. To make her argument, Alombert begins with the Platonic anxiety around writing explored in the Phaedrus and moves into the present. What reveals itself in this alternative genealogy is not a cognitive concern but of the νοῦς, nous, a “crisis of (collective) mind” where social, attentional, and psychic energies become maximally exploitable frontiers for value. What is at stake therefore in the AI moment is not only fears of automation and is certainly not transhumanist fantasies of “singularity,” but a further proletarianization of the life of mind. \ To make this argument, Alombert explains through five essays the composition of the collective mind, its relationship to persuasive technologies and generative AI, and the shortcomings of anthropomorphic metaphors regarding AI, all for a pharmacological critique of AI, that is on the basis of its effects.
In the first essay, Alombert details her definition of “mind,” in so doing critiquing a particular cognitive account on which contemporary AI ideology rests. For her, to say “the” mind at all is a misnomer. As Alombert writes, “The mind is neither a material nor immaterial thing, neither a biological thing nor a mechanical thing, because the mind is simply not a thing at all.”[6] As such, the activities of mind do not consist of the material edifices of the brain but rather circulations between individuals, “techno-symbolic milieus” that includes language but also information technologies dedicated to mind’s direct targeting. Alombert here borrows from B.J. Fogg in designating such supports as “persuasive technologies.” Furthermore, taking from Stiegler, the process of graduating technical supports, from writing to today’s networks, involves an intensifying externalization of the processes of the mind. This is not to say that there existed a golden age of human existence before technical supports, because the externalization of minds also necessarily implies on the other end a psychical internalization. This tends to take the form of something like recognition, where a mass of information becomes discernable in discrete entities. In the AI boom, the interest therefore becomes clear: a different externalization of mind, and indeed further reterritorialization of the mind in technical supports, such that a new frontier emerges near epiphanically in the attention economy.
While the second essay builds off the first, it also introduces two distinctive dimensions to Alombert’s argument: digital governmentality and the failure of anthropocentric metaphors to describe AI machines. Alombert describes neoliberal “nudge theory,” the sense that in an environment where individuals cannot make optimal decisions, they ought to be influenced directly by their environment. In digital infrastructures, this penetration of everyday life is even more total. While Alombert does not mention it, the desire for a “science of government” that calibrates the uncertainty of political actors was frequently thought in lock-step with persuasive technologies. “Father of American journalism” Walter Lippman opens his treatise Public Opinion with “The World Outside and the Pictures In Our Heads,” and the arc of his thesis is that journalism must intervene to correct for subjects gone awry. The wish that AI fulfills is therefore according to Alombert a “neuropower” which might be able to calibrate for the uncertainty of minds. However, this ideology is based in part on a mistaken extended metaphoricity that pervades AI discourse, namely the immediate equivalence of human thought with machine activity. This is problematic empirically for a number of reasons, but Alombert offers the further motivating factor on the part of neuropower that machines able to exactly explain phenomena would do away with the necessity of interpretation and novelty. Interpretation, contemplative thinking, these activities of mind open space for resistance that capitalists and AI evangelists desire to discard. These pharmacological perspectives open up a further reconsideration of the relationship between human and machine, not metaphorical nor analogical, but rather “organological,” taking from Georges Canguilhem, a perspective that instead privileges the relationship and concretized systems of techno-symbolic milieus.
For the third essay, Alombert takes a critical look at the operation of generative AI technologies in the present. Starting with a further detailing of the insufficiency of metaphor in apprehending human-machine relationships, she goes on to outline the concomitant exploitation of cultural and symbolic resources committed by AI in the present, along with its exploitation of natural resources. That is not to say Alombert ignores the question of energy expenditure. She makes the point that the mystified nature of digital technologies—many are probably unaware of the physical infrastructure necessary for the everyday functioning of the internet, between cables in the ground and satellites orbiting the Earth—are only intensified in the case of AI, where the helpful user interface abstracts the massive cost of the technology. Beyond natural resources, AI also fetishizes the human labor necessary for its everyday operation. Alombert mentions in particular the Kenyan workers tasked with filtering through mass quantities of information at less than 2$ per hour.[7] While these are all important critical considerations, Alombert also details the transformation in cerebral organization implied by the new era of AI. As she writes, “When we ask ChatGPT to automatically generate a text, we delegate to it not only our memory faculties,[8] but also our faculties of synthesis, reflection and imagination.”[9] The automatization of these processes risk, following Stiegler, “a generalized proletarianization” or a subsumption even of those diagrams of intimate, sensuous thought. Alombert also provides a political project here, one that focuses on harnessing AI to encourage instead of routinizing processes of mind, such that AI cannot induce us into a shared forgetting how to think.
The fourth essay works well as a synthesis of the work in the prior three, and in some ways works also as a conclusion to the collection before the epilogue focusing on Canguilhem and Stiegler. First, we ought to escape the binary of technical fetishization on the one hand and luddite-ism on the other. Instead, a focus on the effects of AI, the pharmacological and organological approach, highlights the cost of the automatization of collective mind. But this mind does not relate to the material brain, which is only a node in mind, or alternatively a discourse about the soul. Instead, mind is a kind of circulatory process with certain faculties. The ones most under threat by AI are the expressive capacities of mind: an uncertain relationship to language, to the other, and to the social circumstance. These costs are political, a regime of governmentality that promises to cut human uncertainty out entirely—failing that declaring victory anyway and crushing every exception—but also existential. Those processes of mind meant to engage the biological organism with its world become another value proposition. This is not to repeat the Heideggerian Eigentlichkeit (Authenticity) but to instead highlight the cost of “liberation by AI.” What is at stake is invention itself. There is no actuality to imagining if things cannot be otherwise. Whether it is possible for AI to achieve this is one problem, but another, perhaps more pressing, is the way in which capitalists and by extension capital enthusiastically latch onto this possibility as the next great value source in the attention economy. It is this ideology which is least unlikely to go away any time soon, and which AI only encourages.
The coda to the collection concerns how Canguilhem and Stiegler considered the problem of the technical object. In Canguilhem, Alombert first describes the former’s critique of Descartes, namely how Descartes opened up the possibility for a consideration of human life on the basis of machinic objects. The body and its organs as a machine, for instance. What is lost in this account is what is specific both to the machine as well as the human organism. In this case, Alombert highlights the necessity of tool-existence in the biological organism. As such, technics is not separate from biological life, but rather an extension of it. It is here where Canguilhem derives his “organology” which Alombert follows. Stiegler also follows the organological consideration of technics. For him, so Alombert argues, artificial life is an extension of exosomatic organs, or technical prostheses that can be extended beyond simple tools and expanded to vast apparatuses and networks. This is not to subordinate technics to the biological, the techne to the episteme, but rather to envelop the technical in a process of life such that, according to Alombert, “What we on occasion consider to be the ‘properties’ of humanity, will have to be approached from an organological point of view: not as immaterial processes unfolding ‘inside the heads’ of individuals, nor as cerebral mechanisms unfolding in their brains, but as processes of symbolic exchange.”[10]
What does schizophrenia do? Deleuze and Guattari famously argue that “A schizophrenic out for a walk is a better model than a neurotic lying on the analyst’s couch.”[11] For them, what a schizophrenic unfolding revealed was an entire sphere of becoming that existed outside, could not be rendered legible to, the familiar structures of the oedipal triangle. The digital schizophrenia of today, which for Alombert is the split in sentiment regarding AI—on the one hand its enthusiastic acolytes and, on the other, those concerned about its potential to destroy the world—also indicates something similar. Not because humanity is on the edge of bringing about Skynet, or VALIS, but rather because the AI moment might be able to demystify mind and the activities meant to belong to it. The reversal here being that in a moment where the cognitive model embraced an expansion past its utility, an alternative description of mind may demonstrate itself. Alombert therefore finally rearticulates, but this time more forcefully and more exactly, a theme familiar to discourse about living. Namely, the necessity for slow life in the fast one, which in Alombert’s approximation, cannot be achieved alone.
[1]Verma, Pranshu and Tan, Shelley. “A bottle of water per email: the hidden environmental costs of using AI chatbots.” https://www.washingtonpost.com/technology/2024/09/18/energy-ai-use-electricity-water-data-centers/
[2] Fitch, Asa and Gottfried, Miriam. “How Wall Street Lenders Are Betting Big on the AI Boom.” https://www.wsj.com/tech/ai/how-wall-street-lenders-are-betting-big-on-the-ai-boom-25b38259.
[3] Simondon, Gilbert (Trans. Adkins, Taylor). On the Mode of Existence of Technical Objects, p.134. (Minnesota, Univocal, 2017.)
[4] Alombert, Anne (Trans. Ross, Daniel). Digital Schizophrenia & Other Essays, p.5. (K. Verlag, Pensées soignées, 2025.)
[5] Ibid. 5
[6]Ibid. 14
[7] Perrigo, Billy. “Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic.” https://time.com/6247678/openai-chatgpt-kenya-workers/
[8] This is, as Alombert notes, the Platonic problem.
[9] Ibid. 79
[10]Ibid. 118.
[11] Deleuze, Gilles and Guattari, Felix (Trans. Hurley, Robert et. al). Anti-Oedipus : Capitalism and Schizophrenia, p.9 (Minnesota. 1983).
JOURNAL OF THE FIRST OUTERNATIONAL 