Sunteți pe pagina 1din 2

Algorithmic governmentality: a passion for the real and the exhaustion of the virtual.

Antoinette Rouvroy
By algorithmic governmentality I mean an unprecedented mode of government fuelled mostly
with infra-personal, meaningless but quantifiable signals (raw data and metadata), addressing
individuals through their profiles behavioural patterns produced on a purely inductive base
rather than through their understanding and will. Presented as personalization (in marketing
applications), augmented reality (in entertainment and self quantification contexts) or routinized
metabolism of security (in law enforcement, prvention of crime and terrorism, fraud detection,
employment screening), data mining and profiling techniques seduce industries and governmental
institutions with promises of real time, automatic, and thus allegedly objective detection,
sorting and forward looking evaluation of the invisible opportunities and risks carried by
individuals. Opening the way to pre-emptive action to secure commercial profit and forestall
dangerous or sub-optimal behaviors, the atunement of individuals (informational or physical)
environments and interactions according to their constantly evolving profiles is an
unprecedented mode of government presenting at least the two following aspects:
1 Immanence: a passion for the real.
The technical ideology of Big Data 1 is not a cause but a symptom of a passion for the real - not
for the realism of representation, but the real as such - the ambition of reaching the world
without any kind of (institutional, political, symbolic, linguistic,) mediation or intercession
(except virtually transparent, almost naturalised computers), in a perspective emancipated from
the yoke of representation, from all kinds of tests (A. Ronell), from all kinds of - even statistical norms (in datamining and machine learning, even the quantitative notion of average and the
normative figures of the average man or the normal man - disappear). The advent of 'data
behaviorism', appears as a revenge of the plane of immanence, the advent of a body without
organs (G. Deleuze) where the role of the body is eclipsed or taken over by data, in a closure of
the digital upon itself through a recursive loop. In the Big Data ideology, the territory becomes its
own map, a map emancipated from the constraints of the situated point-of-view and from the
rules of perspective, a seismograph anticipating fluctuations in a digital world where sign(al)s
and things have become indistinguishable, and where modelization performs its own designs. To
the crisis of (political, institutional, symbolic,) representations and mediations, algorithmic
governmentality (and Big Data as technical ideology), fed mostly with raw data (infra1 As technical ideology (unlike political ideology la Althusser, technical ideology does not recruit,
engage, interpellate any subject - the sole subjects of algorithmic govern mentality being supra-individual,
impersonal patterns composed of infra-individual data ), Big Data presupposes material procedures of
production (data, even raw data, are never given, they always emerge from sometimes sophisticated
processes of production, which, in the case of raw data, involve procedures of anonymisation,
decontextualisation, disindexation,..), but these material procedures of production are rejected in an
indistinct background as to allow instead for the new data-behaviourism to appear as an overall
interpretation of reality supposed to stand on its own feet, as if it was reality as such that spoke in a
completely dematerialised, cloudy voice, cut from the actual practices from which it nevertheless derives.
The Big Data ideology popularises the ideas that raw data are natural , untouched emergences,
that Big Data are exhaustive, that datamining does not involve any kind of selection of datasets, that they
are therefore absolutely objective and non-discriminatory.

personal signals, computable despite carrying no meaning by themselves), and therefore invested
with dreams of absolute immanence, mechanic objectivity and impartiality offer a radical
response: there is nothing left to be re-presented, digital signals and things are indistinguishable
(sign(al)thing is the semiotic disease of Big Data), knowledge does not have to be produced
anymore, everything is always already present, available, latent in datasets, awaiting to be
discovered by the grace of algorithms.
2. Preemption: exhaustion of the virtual.
Algorithmic government has not ethnic impurity, leprosy, pleague, mental disease, as targets
(as biopolitics of populations had, according to Foucault), but radical uncertainty as such, that is,
the conditional mode of what bodies could do . As an immune system of pure actuality,
algorithmic governmentality is above all a strategy of actualization, or an exhaustion (and thus a
neutralization) of the virtual, or the potential. Preemption replaces prevention: the systematic
actualisation and thus neutralisation of what only exists as potentiality or virtuality dispenses
from dealing with the causes of phenomena. The algorithmic government knows (in)dividuals not
as persons endowed with (real or supposed) capabilities of will and understanding, undergoing
pain and pleasure but only as temporary aggregates of infra-personal data exploitable at an
industrial scale. Therefore, the subjects of algorithmic governmentality are not persons but
the infinitely indexable and profileable entities, competing with each-other at a quasi-molecular
scale, affected by alerts and stimuli generating reflex responses in a real-time economy of
reputation, risk and opportunity. Algorithmic government affects individuals by pre-occupying
their future, up to a point where individuals are always already preceded by desires they have
not even had the time to build nor express for themselves.
Algorithmic governmentality, folding the digital upon itself, fantasizes an existence exempted
from the burdening experience of a world in excess of calculation and neutralizes that which,
precisely, could only arise through encounters with what has not been foreseen, nor attuned for
us: thought. Re-claiming a space for critique amounts to deflate the ideology of Big Data
(immanence and the exhaustion of the virtual) by de-naturalizing raw data and datasets, by
allowing into speakability all that is not easily translated into digital data and which does not
count into algorithmic governmentality, by pointing to the preference one may have for making
projects without (too much) predictions rather than making predictions without (any) project.

S-ar putea să vă placă și