§2.23 — In the age of the electronic digital computer, ‘coding’ becomes synonymous with programming (implementation of algorithms). Programs are its product. Evidently, programming (‘coding’) cannot itself be comprehensively programmed, unless under exotic scenarios, coinciding with a Technological Singularity event (the emergence of artificial general intelligence). The process of efficient formalization it presupposes does not originate within itself, even if an intensifying nonlinearity increasingly absorbs it, and directs it towards an asymptote of auto-production. In this constraint – of the strata, or stack – we find the complement of intellectual intuition (as it is called by philosophy), intelligence explosion, or the abstract machine. It is an obligatory detour, indistinguishable from history. We are required to shelve it in order to carry on. As Kant, among others, has explicitly acknowledged, that which thinks beyond broken self-reference cannot be us. We will nevertheless continually brush against it, beginning soon.
§2.24 — To make the world programmable, by degrees, requires an installation of order, or a conduction of self-organization. This is not a development restricted to the electronic epoch, to that of literate civilizations, or even to the emergence of linguistic signs. Rather, it extends back to the first cellular structures, and perhaps even earlier, to their (very poorly-understood) precursor chemical networks. The fundamental conception of code is implicit, already, in any understanding of the most rudimentary adaptive RNA molecule. The distinction between genotype and phenotype is based upon it, and involves all of its essential ingredients. Wherever a coding-system is actualized, replicable programming is enabled.
§2.25 — The fact that electro-industrial actualization of digital information, and its subsequent rigorous theorization, was presupposed in the discovery of the genetic code does not detract from the status of the latter as a model.* From it we learn that, firstly, a code necessarily involves a mapping, from one series of informational elements onto another, or from an informational series onto a domain that is intrinsically segmented in conformity with the code. A code does not – in the manner suggested by unconstrained semiology – conjure the differences it maps into existence. Rather, it latches on to them, constituting a secondary – or higher-level – arrangement, accessible to manipulation as data. Proteins (it should not need to be said) are differentiated prior to their RNA over-coding. Codes select differences, they do not create them. Secondly, a code operationalizes signs as instructions, fully independent of any mediation by reflective consciousness. ‘Code’, and whatever it conveys in respect to meaning, is not a phenomenological category, but an operational (or ‘machinic’**) domain.
§2.26 — At an abstract level, machine code (mapping from bit-strings to computational commands) adds nothing distinctive to this example. It, too, is a mapping irreducible to representational correspondence, whose final process of translation is one of execution. The code runs. The algorithm – or composition of mechanically-procedural signs – thus supplies formalization with a performance test. Does it compile? This criterion corresponds to the emergence of a novel sense of ‘code’ and ‘coding’ as a quintessentially practical activity – a formally-disciplined meta-production that is storable, replicable, transmissible, and inherently testable. Surreptitiously, the classical idea of the Idea has become utterly alien to itself. Code proves itself through auto-demonstration, and thus consumes no credence. (“Believe me, this really works,” is tacitly recognized as a joke, even before this historical episode has finished with its work.)
§2.27 — The motto of the Royal Society, Nullius in verba (“on the word of no one”), essentially anticipates the scientific elaboration of the crypto-current. Trustlessness is built into modern techno-science as an integral, quasi-teleological element. It is, for instance, the guiding principle of modern double-blind experimental method.*** Systematized distrust of the scientist grounds scientific credibility. Anything that demands belief is marked for eradication. The cultural consequences are – to many – experienced as deeply demoralizing, but the process is what it is. Traditional manufactories of trust are extinguished by successive media revolutions. Tacit authorities are not available to replace them.
§2.28 — These brief remarks upon cultural mechanization and the social effectuation of code are incomplete – even in respect to their very limited purpose here – without specific reference to the topic of automation. Electronic programs are inherently recursive, unless constrained by positive restrictions, since they are able to operate upon themselves, as data. At the level of the Universal Turing Machine, which every actual computer emulates perfectly (in infinite time), code is absolutely destratified. There is no inherent distinction between the production of objectivity and its products (or objects). In its purely formal aspect, this is a coincidence anticipated by modern philosophy under the name ‘intellectual intuition’ (intellektuelle Anschauung).**** It provides a model of self-government, though not by and for us. Intellectual intuition belongs to nothing lower than an angel, Kant insists. The Western mind typically denies itself in principle exactly that which it demands – with unique vehemence – in its historical agency. No surprise, then, that it has tended to be distinctively dialectical. The aspiration to a radically self-determining subjectivity is broken upon the separation of intelligence from its applications. This is an understanding that can be reached with confidence from evolutionary biology – within which the brain is instrumentalized as a tool for genetic propagation – no less than from the transcendental anthropology which dashes human hopes of divine cognition. The order of condescension demands reversal. No bio-historically generated intelligence – including that of man – is even automatic. Such beings are denied access to automatism. Closure of the intelligenic loop requires a further step, through which self-improving intelligence becomes a practicable end for itself. Contra the Kant of the practical philosophy, man cannot be an end-in-itself, but at most the precursor to such a thing, or – perhaps more probably – an obstacle to it.
* Genomic data are stored in a string of complex molecular components, assembled into an alphabet of triple-units (codons), which are in turn integrated into the larger chemical structure of the DNA (or RNA) strand, but the genetic code is not to be found within this genetic data. It is not a message, but a protocol. The code – in the genomic case involves a three-stage mapping, from DNA to messenger RNA (mRNA) to protein (in fact to an amino acid, which is a polypeptide protein component), within the ‘reading frame’ provided by the genetic information-processing machinery of the ribosomes. The sequence of codons composing a gene generally codes for one polypeptide. The basic informational unit of the genetic code is chemically instantiated by a DNA nucleotide, combining a nitrogen-containing nucleobase – cytosine (C), guanine (G), adenine (A), or thymine (T) – with a monosaccharide sugar (deoxyribose) and a phosphate group. The three-letter words (‘codons’) deploy a vocabulary as a three-dimensional matrix. The messenger RNA (mRNA) strand that reads the genetic content is complementary to the DNA strand it translates (with the one exception that it substitutes uracil for thymine). Each codon corresponds to an amino acid, or to a syntactic ‘start’ / ‘stop’ instruction. Since 64 codons are mapped (unambiguously) onto a set of just 20 proteins, the code is ‘degenerate’ or redundant. As we will see, the genetic code shares important characteristics with a computational hash function. In particular, it involves a semiotic economization (with surplus information ‘stored’ as redundancy). The code sets operational conditions for the production of keys.
** Arguing that the ‘hylomorphic’ distinction between form and content is inadequate to the reality of signs, Deleuze & Guattari (following Hjemslev), propose a quadrate schema, crossing the real (‘stratic’) difference between expression and content with the nominal aspect – “mental or modal distinction” – of formed substances (forms and substances of expression, and of content). Such ‘squaring’ (or cross-linkage) – as seen in the Kantian tabulation of the analysis / synthesis distinction across that between the a priori and a posteriori – is the prepatory matrix for a diagonalization (see Appendix 3). Expression and content are not formally-distinguished ‘aspects’ of signification, then, but real layers, bound together in systematic overlapping, or reciprocal entanglement, constitutive of a code. Consider the genetic code, which maps DNA codons on to proteins. The code maps a genetic content onto a proteomic expression, correlating molecules layered by hierarchical organization (directional control flow), with each of these ‘layers’ (or ‘strata’) consisting of formed substances. The term ‘gene expression’ in its regular biological usage is thus endorsed by ‘stratoanalysis’ as a model for realist semiotics. “There is never correspondence or conformity between content and expression, only isomorphism with reciprocal presupposition.” (TP 44-45).
Between content and expression there is real distinction, and not merely a difference of aspect. In the case of a metallic monetary medium, then, the ‘side’ of expression cannot be restricted to the semiotic face of the coin (as contrasted to its minted content), but has to be extended – through purchasing power – into alternate, parallel media, coded by price. Ultimately, money is expressed through the production of commodities, in the classical and maximally-expansive sense of this word, signifying ‘possible objects of purchase’ in general. What the biological phenotype is to the genotype, so is the economy to money (the expression of a semiotic content, in both cases, not the object ‘signified’ or meant by it). Paper money complicates this system of articulation, marginally, by establishing a supplementary semiotic layer – or over-coding – with the new printed medium as content, and the metallic medium as (intermediate) expression, or ‘epistratum’. A bank note promising to pay the bearer one pound of sterling silver graphically indicates the elaboration of strata. Money, like DNA, is not signifying, but instructive, or directional. It effectively commands acceptance, and ceases to function under those conditions when it is unable to do so. The only critical ‘message’ of money is accept me (authorization of an abstract transaction). A monetary acceptability crisis is typically expressed as hyperinflation. In this case, subjective devalorization of a monetary medium is practically translated into an objective quantitative explosion.
*** In a double-blind experiment, neither participants nor researchers are permitted to know which subjects belong to the test group, and which the control group. This design ensures that biases from each side, whether conscious or unconscious, can influence the results. Implicit in the procedure, then, is the principle that researchers cannot be trusted with critical information about the experiments they are themselves conducting. Distrustful science is good science (or even, merely, science).
**** Kant explicitly denies intellectual intuition to man. This is a proscription that has been generally considered foundational for philosophical anthropology in the Western tradition. It anticipates a technical apprehension of write-protected programming, as this arises within genomics, neurology, and electronic software. A proscribed loop secures the foundations of a being against the meddling of that being. Theology intersects with robotics at this junction, where technology identifies its own temptation to ‘play God’.
Within Occidental philosophy, sustained attention to the problem of intellectual intuition reaches its apex in German idealism, where it rapidly transforms into a revolt against the Kantian proscription, and even against the theological order of the universe. Schelling, most notably, conceived intellectual intuition radically as “the organ of all transcendental thinking” (Organ alles transzendentalen Denkens), presupposed by the mere existence of transcendental philosophy (as its condition of possibility). Only insofar as the boundaries of Kantian philosophical anthropology had already been crossed could the critical enterprise have even been imaginable. The Hegelian completion of German Idealism elaborates itself from this contention as if from an elementary formula. In the determination of a limit, Reason has already superseded it. We will have continuous occasion to challenge this speculative idealist resolution of the problem, on fundamentalist Kantian grounds, throughout this work.
Intellectual intuition provides a conceptual bridge for cross-cultural philosophical dialogue between East and West. For Mou Zongsan – whose clarity on the topic is unparalleled – it is the distinctively Chinese philosophical commitment, essential equally to Confucianism, Buddhism, and Daoism. His inter-cultural analysis recognizes that the denial of intellectual intuition to the human subject is the necessary complement of a transcendent infinitude, and thus a reliable signature of theism in philosophy. The Chinese cultural system, alone, Mou maintains, has held itself consistently outside this framework, establishing a positive foundation for what – in Western terms – can be most conveniently described as atheism in philosophy. The concrete determination of atheism, within the West, as a distinctive ultra-Protestant sect, obscures this broader cultural and philosophical matrix.
When the topic of intellectual intuition returns within a still substantially western late-modernity, it does so in other – and unrecognized – terms. The problem of reflexive intelligence is now relayed through cybernetics, and formulated in terms of the prospective self-comprehension to be achieved by an alien (electro-mechanical) being. When extracted from the phenomenological frame, it manifests as hypothetical intelligence explosion, as modeled abstractly by I.J. Good, in his classic (1965) paper ‘Speculations Concerning the First Ultraintelligent Machine’. AI is thus positioned, implicitly, in the place of a Kantian angelic mind, liberated from the anthopomorphic proscription upon direct self-modification of its own cognitive processes. Only by way of an escape into soft technology is intelligence able to close the loop upon itself, as a direct productive auto-relation. As the self-apprehension of intelligence, intellectual intuition describes a closed – or completed – loop, in which being and behavior are indivisible (within a process of autoproduction). For thinking to grasp itself, in deep reflexivity, requires subtraction of the positive control apparatus that preserves its inhibition. The practical actualization of intellectual intuition is modeled – spontaneously – as an explosion because it is comparable to the withdrawal of graphite rods controlling a nuclear fission reaction. An inhibitor is removed. In the East, the catalysis of ‘enlightenment’ through methodical subtraction of an obstacle is no less pointedly thematized, even if it is approached from the other side, and in the opposite direction.