An Algorithmic Virtual, Also Embodied

The novelty that can emerge during the exclusive execution of the algorithm – associated with a given dataset –, such as the generation of new, unpredictable and surprising images, also passes through the artist’s body. It owes its possibility, as novelty, to the artist’s affective contact with the images, from which a surplus of meaning and sensation emerges, for, from a structural point of view, the calculated image – indeed – already possessed an intelligible consistency, even before it had been materialized.

However, before the image can enter into affective contact with the artist, two prior transformations must take place at the level of the algorithmic core, transformations that are far from trivial. First of all, the image acquires a topological structure once the calculation is complete¹. Before that, it was contained only potentially in the algorithm and its data. Following the calculation, the image can be “viewed” (on a screen, for example), and only at that moment does it adopt a concrete spatialized structure. Thus, viewing this image on a small screen will not be equivalent to seeing it projected in large format, with this or that color profile, seated in a movie theater or strolling through an exhibition hall, etc.

This form of the virtual, which takes shape during the execution of the algorithm by way of the artist’s experience, we will call the algorithmic virtual. This virtual manifests itself first in its topological form, then in its spatial staging during viewing; but above all, it becomes effective on the symbolic and sensible level, through affective contact with the artist. The algorithmic virtual differs from the analog virtual in that the latter is actualized in an exclusively divergent manner, whereas the former first passes through two transformations in which a form of resemblance and limitation is expressed: the one induced by calculation², which gives the image its topological structure, then the one linked to its spatialization (rendering, viewing), during which it takes shape in space. The algorithmic virtual envisaged here would indeed be comparable to the possible – as understood by Deleuze, Massumi and Lévy, and which is realized in a non-productive way – but only if one disregards the final stage of actualization: the affective reception on the part of the artist, from which novelty arises. Yet this stage is fundamental.

It remains to be emphasized – once again – that these forms of the virtual, the algorithmic virtual as much as the virtual tout court (for us, the analog virtual), could, within the framework of this research, constitute themselves only as immanent entities. In this sense, these forms of the virtual naturally fit into that fundamental, metaphysical level to which the works considered in the central corpus have led us: a primitive and preindividual stock, where gesture and affect are already at stake; an undifferentiated fundamental plane, which could only be material, and where dynamics are expressed that, through their inaccessibility, engage new forms of mysticism: immanent mysticisms.

1. That is, the image, not yet spatialized and thanks to the calculation already performed, now possesses a description in terms of I and J for the positions of the pixels, each pixel having an RGB color component.↩︎

2. Beatrice Fazi suggests that there is a form of indeterminacy within the very formal logic of computation, which leads her to propose the idea of a “contingent computation”. The point would be to think an aesthetics proper to the digital, to the discrete, founded on the incomputability formulated by Turing and on the incompleteness of formal systems demonstrated by Kurt Gödel in 1931. This research has studied Fazi’s proposal with interest, without, however, managing to date to grasp the consistency of an aesthetics outside the sensible, outside affect. See on this subject: M. Beatrice Fazi, Contingent computation: Abstraction, experience, and indeterminacy in computational aesthetics. Lanham : Rowman & Littlefield, 2018.↩︎