Even if the question of the symmetry of time is still a field of debate in the philosophy of science, it is undeniable that, if we agree with the probabilistic monism of contemporary science, and posit that the universe is time-symmetric or that it is temporally reversible, then it becomes difficult to explain temporally-directional phenomena like the arrow of time in thermodynamics, or the outward propagation of the wave, without recourse to some notion of subjectivity. We are to understand that if we see the universe we see, as irreversibly moving in the direction of higher entropy, it is because we are oriented within it: we are negentropic organisms, and thus pointed in the direction of the universe’s expanding and cooling. Our being-toward-death, is perhaps nothing more than a reflection of this negentropic resistance to entropy. The way contemporary physics sees it, this orientation in time corresponds to our asymmetrical selection of the universe due to our thermodynamically oriented nature, whereas “objectively”, time is perfectly symmetrical and all processes are reversible.
Now let us recall the argument against the absolutization of totality, articulated on the theory of transfinite sets, that Meillassoux (after Badiou) develops to defend his speculations on the necessity of contingency. His argument is that it is impossible to totalize the all, for according to Cantor’s set theory, the set, by definition, contains itself, its full set, as a subset of itself, in addition to its other subsets. This means that the set of all sets contains more than it contains, which is evidently paradoxical. We thus imagine an infinite series of inter-embedded sets of increasing cardinality, the set of all sets perpetually being surpassed by its contents, outnumbered by its parts. Moreover, the impossibility of totalizing the set of all sets poses an obvious problem for any cosmology that defines the universe as time symmetric. For the time symmetry of the universe rests on the possibility of a temporal objectivity, in the sense of a point of view that sees all, or in other words, a set of all sets. This suggests that we should question the plausibility of the current posture in cosmology, that insists that the asymmetry we observe is illusory in relation to symmetrical reality, since in order to be symmetric, it must also contain itself as an objective totality.
It is only from the subjective point of view that there is asymmetry, and only from an absolutely objective non-point of view that there is pure symmetry. But it is my contention that, following the paradoxes of transfinite sets, both of these poles reveal themselves as untenable. Meillassoux is correct in that there can be no pure objectivity, for the set of all sets cannot be totalized. But what Meillassoux misses is that this argument holds for all sets, not just the set of all sets. So it also means that there can be no pure subjectivity, no pure closedness in the first place, because by definition, subjectivity, as a subset of the whole, also cannot contain itself. It is not closed: it escapes itself through its parts, bootstraps itself out of itself. Subjectivity as asymmetry presupposes the correlationist exclusion and veiling he opposes. Paradoxically, it is the same self-veiling nature of (self) observation, which Meillassoux opposes in the individual (and which he reduces to linguistic correlationism or idealist solipsism) that allows in the first place for his argument against the totalization of the set of all sets, for if the whole can never be totalized it is because “it must first cut itself up into at least one state which sees, and at least one other state which is seen. […] In this condition it will always partially elude itself.”
The universe can thus not be objective, for objectivity is outside of time (which is subjective) and can thus not account for the contingent event. But there is a sense in which Meillassoux’s argument for a hyperchaotic universe leads him to defend a position as untenable as that of the solipsist. For now hyperchaos can thus be applied nowhere and to nothing: it does not allow for the actualization of the event, for no set can contain itself enough to be affected by an event, no set can loose itself to the event for it has nothing to loose in the first place, or more precisely, it is always already lost to a single event of self observation. As Deleuze would say, nothing in this paradigm becomes the event’s quasi-cause, for there is nothing for the event to happen to. Indeed, hyperchaos is the inversion of soplipsism: to the berkleyian idealist solipsism of an anti-symmetry and absolute closedness, where nothing really happens, hyperchaos substitutes the absolute symmetry of the non-totalizable openness of the non-all, where there is nothing to happen to.
But the pharmakon seems to point us toward a third alternative, one that allows us to understand the world as a hybrid of closure and openness, between hyper solipsism and hyperchaos. Objectivity and subjectivity only obscure the depths of the strange attractors that tessellate the pre-individual field. We should avoid reducing everything to one or the other, to the marked or unmarked, for it is what happens in the interstice that counts : there is all the richness, the complexity, the consistency of the world.
So, to the open and hyper chaotic universe of Meillassoux, and to Luhmann’s universe of inter-embedded closures, I believe we must oppose the universe that Deleuze, after Joyce, called a chaosmos. A universe that is intrinsically hybrid, that is stratified between chaos and order, between openness and closure. It becomes equally nonsensical to consider time as a process that passes from the past to the future. Indeed, according to Deleuze, time can only pass from the virtual to the actual, and this, by traversing the intensity of the plane of consistency. All the possible results of the dichotomy of self-other, subject object, interior-exterior, remedy-poison, dissolve into the consistency of the process of actualization, just as the inside and outside unfold out of the pharmakon. And the mistake is to always try to reduce everything to one of the many attractors in the chaosmos. The universe is hybrid, and the pharmakon is its consistency.
The Big Bang might be construed as the fixed point attractor of objectivity. Like the first distinction, or the arche-trace, it is a substantive mark, and can only supplement actualization. It holds all of becoming in a single point, outside of time (the cosmological singularity). But each moment of self-reference produces its own intensive singularity that bifurcates from the spatial plane and produces its independent temporal axis. There is therefore, strictly speaking, no Big bang, for all the little bangs that compose it resist their totalization. The universe would be better understood as a web of strange attractors, chaotic attractors that have as their first characteristic to be withdrawn and plunged into the infinite regress toward a multiplicity of initial conditions.
Let us take one last example before we conclude: according to the discoveries of the theory of complexity, such as in the work of Stuart Kauffman, self-organising systems seem particularly attracted to the edge of chaos. A self-organizing system is not simply selective and retentive, it is not simply negentropic: in order to be efficient, it must situate itself on the edge of chaos. A system that is too ordered remains fixed and crystallized, incapable of evolving, of varying, of creating new forms that can survive in the changing environment. By contrast, a system that is too chaotic cannot hold enough order to contain itself, not enough structure to inscribe a memory, a logic, a program. But between these two poles, there exists as third regime: where order is liberated from the fixed point attractor and rises into differentiation, but also where chaos allows itself just enough body to not evaporate and loose itself. Christopher Langton, in the 90’s, even discovered that the universal computer potentially emerged in the fine layer that separates the regime of order from the regime of chaos. For it is there, between order and chaos, that virtual computing can spontaneously occur, as described by Alan Turing in 1936: a virtual machine that would read both the description of the machine it simulates and the data it computes from the same series of variations inscribed on the edge of chaos.
So if we grant the edge of chaos this potential capacity to virtualise autonomous universes, to simulate them, as does Turing’s machine, then the self-organizing universe escapes itself perpetually in its race toward higher entropy. The edge of chaos is thick, for if the virtual is reborn on this edge, if little universes are simulated somewhere between the birth and the heat death of the universe, it also means that perhaps, somewhere on the path to higher entropy, the universe reaches a zone where the system’s behavior bifurcates into heterogeneous dimensions. This recalls physicist Lee Smolin’s theory of the fecund universe. He speculates that each black hole in the universe produces its own Big Bang behind its event horizon, with its own forces and constants. Each universe only produces offspring (little bangs) to the extent that it produces black holes. The black hole acts upon the system of the universe, somewhat like the self-referential statement acts upon the mathematical system. Smolin’s black hole is the pharmakon of the universe, its exteriorization, its escape, its flight from self, just as Gödel’s self-referential statement is the pharmakon of mathematics. The pharmacological is composed of this fractal-like web of inter-embedded holes and strange attractors that weave themselves around each process of individuation and constitute its preindividual field.
The pharmakon in this sense relates to the concept of the ()hole complex in Negarestani’s Cyclonopedia, where each whole escapes itself through its own self-referential (black) holes, that form where cyclones and spirals conspire and become complicit, sharing inward folds and spires, and perpetually defer totalization. The point here is that the pharmakon should no longer be reduced to a mere undecidability between outside and inside, between poison and remedy, for it is rather the consistency of individuation itself. Absolute closure and absolute openness are nowhere to be found. There are only pre-individual individuals, there are only partly discrete identities, there are only partly continuous continuums, for the fabric of the universe is intrinsic hybridity, and the body of becoming is composed not of organs, but of holes, escape tunnels and lines of flight.
 I agree here with Peter Hallward’s argument against Meillassoux :
Peter Hallward. Anything is Possible: A Reading of Quentin Meillassoux’s After Finitude dans Bryant, L. R., Srnicek, N., & Harman, G. (2011). The speculative turn?: continental materialism and realism. Melbourne, Victoria, S. Aust.: re.press. p. 130-141
 Against the predominiance of the closed system in second order cybernetics, Mark Hansen has proposed an interesting alternative, with which the pharmakon aligns perfectly, that he calls the system environment hybrid, drawing from, among others, the Simondon’s individuation and Guattari’s notion of the machine.
Mark B. H. Hansen, System-Environment Hybrids dans Clarke, B., & Hansen, M. B. N. (Eds.). (2009). Emergence and Embodiment: New Essays on Second-Order Systems Theory. Duke University Press Books.
 Kauffman, S. (1996). At Home in the Universe: The Search for the Laws of Self-Organization and Complexity (1st ed.). Oxford University Press, USA.