Vincenzo Maria Di Mino - Towards Reza Negarestani, Intelligence and Spirit
Reza Negarestani, Intelligence and Spirit (Falmouth: Urbanomic, 2018)
Vincenzo Maria Di Mino
Bionote: Vincenzo Maria Di Mino (1987) is an independent researcher on political and social theory
Reza Negarestani's theoretical work, within the contemporary philosophical discourse, stands out for its originality and innovative approach to classical and fundamental themes of the Western philosophical canon. The thread that binds the Iranian author's various works is the search for a “philosophy of the outside,” a different archè, on which to build new forms of rationality and new alliances between living beings and other forms of life. Indeed, Cyclonopedia, an earlier work by the philosopher, highlighted the centrality of these themes, setting up the narrative of Reason as an alien element, through a kind of cosmogony populated by monsters, demons and war machines. In his new work, entitled Intelligence and Spirit, the philosopher continues his research into the alternative dimensions of reason, developing it in a comprehensive and systematic way.
Intelligence and Spirit, in the classical philosophical canon, are the elements that characterize the development of rational subjectivity, embedded, in turn, within specific temporal, spatial and intersubjective dialectics: in this case, the fundamental objective of the philosopher's theoretical machine is the elaboration of a general program for artificial intelligence, through the history of philosophy. The philosophical argumentation of the text, in fact, is constructed by the author along two fundamental lines, namely German Idealism and Anglo-American analytical philosophy, up to its contemporary developments. Through the first pole, Negarestani, drawing heavily from Kant and Hegel, further develops the concepts of Geist and self-consciousness, projecting them into the dimension of computation, and, from the second analytical pole, borrows the logical and mathematical rigour of argumentation, as well as the formal dimension of models and structural matrices. By connecting these different analytical approaches, the philosopher aspires to the elaboration of a universal rationalism, whose transcendent dimension is the expression of a universalist tension, capable of materially realizing the egalitarian and emancipatory dimension of the “Intellect.”
The eight chapters of the book (including a substantial appendix that further develops some formal models analyzed above), can be read through four argumentative blocks: the first deals with the theoretical basis of the construction of Artificial General Intelligence (which will be referred to throughout the text as AGI), the second with the articulations of intelligence through the dimensions of thought (time, space), and the third with the social and communicative dimension of this intelligence. The last block attempts to summarize these design hypotheses through the exposition of a philosophy of artificial intelligence, namely the necessary relationship between collective intellect and technological automation.
The first two chapters focus on the analysis of self-consciousness and its projections into the abstract universe of the artificial. From this starting point, Kantian and Hegelian philosophy are analytical machines capable of destroying what the author calls the “myth of the given,” i.e., the empirical dimension that reduces the thinkable to the thought, the subject to the object. Kantian philosophy of knowledge, which places cognitive forms in a transcendent dimension, and Hegelian philosophy, which brings the previous theoretical structure back into the dialectical contradictions of reality, tend to construct the “labour of mind” as an exercise in rationalizing formal intuitions within the historical and spatial coordinates of the action(s) of subjectivity. In this sense, the Geist, as a mode proper to rational agency, synthesizes formal apperceptions at a higher level, and prepares their social and material constitution, acting as a model of interaction of a multi-agent system, and coordinating at a logical level both the rationality of the systemic space and the rational coordinates of each individual actor (55). The Geist, then, is the first artificial step of intelligence, which encodes and structures social interactions internally, and whose intelligibility is given by the continuous integration with a mind-independent reality, which must be constantly explored and made communicable. For Negarestani, the constitutive function of language, both in terms of its subjective dimension and of its productive effects in the social sphere, requires its own model, which he constructs by integrating three different elements: artificial speech, artificial intelligence, and transcendental psychology (89). The dialectic between the three elements emphasizes the central role of language as a mediating element between the different dimensions, requiring ever higher levels of synthesis. This, in fact, allows him to lay the foundations for the definition of the AGI as an element able to constantly generate practical-cognitive skills, to be aware of history and to be able to produce normative judgements (93). In this sense, Negarestani’s theoretical operation takes an opposite direction to that of the theorists who seek to overcome humanity, questioning the simple equivalence between human subjectivities and rational machines: it is not a question of irreversibility, but of constant interaction and transitivity between these different forms of intelligence. Indeed, what Negarestani calls “soft” and “hard” variants of parochialism are dismissed as examples of the constant reduction of artificial intelligence to the anthropocentric realm. Negarestani attacks the myth of the immediacy of knowledge, pointing out that every act of knowledge is preceded by a previous act of elaboration; consequently, the labour of intelligence is the labour of suspending immediacy, thus hearing the constant extension of determined negation, of negation as an act of individuation that links primary apperception and transcendental reason (117). The AGI project, according to Negarestani, is thus materialized through a Big Toy named Kanzi, capable of reproducing the zero degree of intellect, so as to discuss the elements in order to expand its faculties. Kanzi is consequently developed through two models: that of Chu Spaces and that of Virtual Machine Functionalism (VMF). The first is the topological space useful to represent the dualistic interactions of the AGI, those between sensing and thinking, and those between the different psychic and physical elaborations of the received data. The second model acts as a topography of the mind, showing the different levels of the causal and logical connections of the virtual machines, whose specificity is the capacity for modelling and emulation. Through this virtual model it is possible to design Kanzi's construction in analogy to that of the pre-linguistic infant, namely the integration of language models as computational codes (139).
The second analytical part of the book begins to elaborate on the intellectual structures of AGI by leaning on Kantian categories, and pushing them towards a more declined reading through the lens of analytic philosophy. Thus, Kanzi is iconographically represented by a Lego model, whose cognitive perception is shaped on Kant’s three syntheses: that of apprehension in intuition, that of reproduction in imagination, and that of recognition in concept (159). In this sense, the passage from sensation, thus the immediate perception, to sensibility, to conceptual elaboration, helps to designate the spatio-temporal location of the AGI, and requires a deepening of computational logic, in order to translate the empyrea of the former to the level of the normative force of the concept. For the author, this transcendental-computational logic coincides with the “liberal functionalism” elaborated by Hilary Putnam, a non-reductivist vision of the relations between the system and the individual. This systemic model, in fact, consists of both conceptual intuitions distributed on the different planes of the structure, and meta-conceptual elaborations that determine the conceptual and systemic elaboration of the same. For the AGI, there is a constant shift from the non-conceptual reception of bare (fuzzy) data and sensations to their processing in objective, spatial terms, limited to the precarious sense of movement it possesses. The construction of self-consciousness, as a primary objective, is a continuous process of intra-systemic differentiation from the space that surrounds it, and from the relationship it has with the ends pursued, through which it is inscribed in the same space. Thus, this limited consciousness of space is essential in producing the fundamental consciousness of time, through the differentiation of temporal sequences. Temporal dynamics, consequently, are the syntheses of individual apperceptions of spatial objects, and of the meta-categorical capacity to represent them in a virtual dimension, as real categories and as theoretical abstractions (191). Consequently, possessing the concept of the succession of things, the machine is able to translate the set of sensations into a network of relations of equivalence, into a set of sequential codes, which, in turn, is necessary to construct memory as a geometric form. Negarestani adopts an analytical posture rooted in pragmatism and in scepticism, questioning opaque representations of objectivity, as opposed to empirical cynicism, a recurring tone in the contemporary world.
Starting from this theme, the author confronts the different conceptions of time and the predictability of events, focusing on Boltzmann's analytic epistemology and Russell's logic; from the former he borrows the concept of the impact of entropy on the description of temporal sequences, and, from the latter, the description of memory-image as a statistical inference, which has linguistic judgement as its counterpart. The representation of the temporal sequence, following these supplements to the Kantian model, must be linked to objective elements, but cannot be reduced to them, because it is a singular expression of a meta-temporal nowhere reality (243). This external reality is none other than the Absolute Spirit of Hegelian memory, which requires a constant process of reduction to the Concept (Begriff), to be identified within the historical totality. It is precisely the faculty of intellectual elaboration and the linguistic-communicative dimension of AGI, hence the construction of inter-subjective reality, that constitutes the theoretical plot of the next part of the book.
The Geist, in fact, is theorized as a multi-agent system, characterized by the linguistic interactions between individual agents and the goal-attainment probabilities of their actions, which necessitates the construction of self-consciousness as itself a necessary stage in the transition from subjective representation to the objective dimension of the social milieu. The AGI as “I”, consequently, is a vector of synchronic translation of meta-conceptual dimensions into the objectivity of the concept, thus a process of formal individuation mediated by cognitive elaboration. This cognitive machine, for Negarestani, determines the link between the object and the norms of knowledge, or the link between objectivity and the thought that determines the apperception of the “I” as a recognition node of the labour of the mind within a space of formal reconnaissance of other agents, both in the synchronic and diachronic dimensions (271).The role of education, of the cultivation of cognitive activities through the enlargement of the space of mutual recognition, is at the heart of this dynamic of emancipation of the AGI from the realm of sensation, therefore of the construction of its will-to-autonomy. The form that pedagogical practices take is that of horizontal modularity, a set of intra-systemic integration processes of the different levels of cognition, useful for perceiving the complexity of the world. This involves the construction of what, following Sellars, are named semantic and syntactic skills. These can, in fact, be considered in terms of a combinatorial calculus of objective thinking (282), which is necessary for the linguistic and computational structuring of objectivity. Both semantic skills, on the qualitative side of expression, and syntactic skills, on the formal side of encoding, cross the cognitive dynamics of AGI, highlighting the need for a specific interactive grammatology. The linguistic dimension of Dasein, at this point, is the constitutive condition of General Intelligence, which proceeds from symbolic interaction to syntax, and, from this, to general semantics. The encoding of these interactive and relational aspects proceeds by sequential regularities, which can also be reproduced as binary strings that enable agents to select, compress and process communicable and exchangeable data within the system (313-315). Consequently, the self-organizing model of the multi-agent system is based on a generative grammar, capable of translating sound and visible sensations into a finite set of sounds and symbols. There are two operations that characterize these processes: iteration and recursivity (as also theorized by Yuk Hui), namely the capacity to generate and recombine elements within a structure (a syntactic structure, for example) and within a historical memory. The combination of both processes, in fact, generates other structures and allows the transition between the different hierarchies of sense production (322-323), in Sellarsian terminology, the passage from the game to the meta-game and vice versa.
Negarestani uses these further elements to analyze language, at this point a combination of logic and computation, as the sole mediator in AGI's world-building processes. In this case, the philosopher's reference is Brandom's pragmatic linguistics, for which the normative dimension of language is inseparable from the interactive dimension, which indeed redefines the basic rules themselves. Sociality, the relationship between agents, is finally assumed as a substantive feature of linguistic practices, as a self-reflexive form of reason, and as a concrete practice of change through interactive games between agents. The coding of game strategies, from initial regulatory patterns, contributes to the construction of new regulatory patterns, which can be imitated (copycatted) or ordered. The formal dimension of language and the practical dimension of experience find a unitary synthesis in the practical performance of the intellect, that is, in action directed towards specific ends, which, as expressions of the temporal Absolute, also serve as premises of the cognitive process itself. What characterizes the artificiality of AGI is its ability to adapt to the new material conditions it faces in different temporal sequences through the plasticity of its linguistic structure.
The final part of the work is dedicated to the exposition of a philosophy of intelligence, hence of a philosophy of praxis, declined to the relationship between human and artificial intelligences. The first aim of this program is to tear apart Western philosophy, so as to make the Outside thinkable, and to proclaim the equality of all minds (in the manner of Rancière). The connection between the two forms of intellect, in fact, goes in the direction of the decolonization of thought, and liberation from exploitation, through the constant broadening of the conditions of the possibility of equality (408-411). Starting from these premises, Negarestani lists the salient points of this new philosophy, which, aiming at total emancipation, is extended beyond existing space and time, and makes them intelligible through their transformation into concrete data. In this regard, philosophy is a continuous exercise in reflection, analysis, and combination of data, as well as the transformation of these elements into normative forms of judgement. At the same time, it assumes negation as a central element of the architecture of the intellect, as an element of differentiation and socialization of cognitive functions. Consequently, this philosophical program is also an operation of re-building, of connecting the universal dimension of the Intelligence with the local dimensions of individual political and social issues (for example). Further, this is a reflection on the realization of all potentially realizable forms of cognition. All of these exercises and proposals redefine the main purpose of the philosopher's reflection, namely the conditions of possibility of the intellect as such, the primary conditions of thought and the logical, historical and social conditions of thought itself. In this sense, nihilism becomes the key to emancipating the intellect from the constraints imposed on it by philosophical reflection. Free from the “myth of correlation,” as Quentin Meillasoux puts it, the author shows how this Geist is timeless, and, consequently, cannot be located in any space-time coordinates (480). Consequently, this intellect, situated in the suspension between the Absolute and contiguity, can finally realise itself as Hegelian Absolute Knowledge, that is, the formal condition of all possible thought (502), in the concrete unity of the world and intelligence. As negation of all previous forms, it stands as a philosophy of the impersonal, serving as a matrix of processes of individuation—of singularity through the universal. In other texts, Negarestani has called this approach “inhumanism,” as a theoretical reflection that breaks out of the stale dialectic of existing parochialisms (mainly discussions on the human\post-human) to amplify the humanist rationalism at its core. This means that Reason can only be realized through the continuous negation of its own presuppositions. The search for the Outside, as a form of estrangement from reality, is veritably the central assumption of this theoretical speculation, which can lead to the main aim of philosophical research, the unveiling of truth through transcendental (ir)rationality.
In conclusion, Negarestani’s dense, thoughtful and compelling book opens the way to new kinds of reflections on the relationships between different forms of intelligence. Firstly, it proposes a mapping of intelligence, overlapping the rational and artificial dimensions, balancing the contribution of classical speculation with analytical rationalism. Secondly, the philosopher's analytical operation, which suspends the construction of the intellect between the abstract and the concrete dimension, can be compared to the philosophy of the trans-individual proposed by Paolo Virno, based on the Marxian concept of General Intellect—the general social knowledge embodied by labour-power. As an impersonal expression of a collective subjectivity, this concept can be linked to the dialectical dimension of the construction of the AGI, as a radically alternative matrix of antagonistic subjectivation. Consequently, the “construction” of the AGI deepens the “accelerationist” discourse on the liberating function of technology, starting from its own unexplored conditions, potentially guaranteeing, through widespread social conflicts, greater social justice. If we want to hypothesize a political translation of Negarestani’s theoretical device, we can say that it aims to construct an alternative to the different forms of governmental algorithmic control and the consequent encodings within archives and databases through the construction of diffuse and self-organising artificial networks, and more egalitarian uses of artificial machines.
By equating the Absolute Spirit with the Supreme Good, Negarestani hides the utopia of liberation between the lines of this work and builds a theoretical instruction manual, metaphorically pushing readers to rethink the very conditions of the production of cognition, and, therefore, of thought as its primary effect. Rethinking the conditions of possibility of the intellect, in this sense, means rethinking both human subjectivity and the ways in which knowledge is produced and socialized. The technological-artificial dimension, under the sign of the “new alliance” with the human, is necessary for planning actions and making purposes predictable, and for allowing the design of concrete alternatives to the collapse of the ecosystem. The multiple dimensions and unexplored potential of hybrid intelligence, freed from the one-dimensional shadow of anthropocentrism, can contribute to the construction of new cosmo-alliances and the search for the good, collective life.
 See: Yuk Hui, Recursivity and Contingency (Roman and Littlefield, 2019).
Published on: June 10, 2022