Suchergebnisse

La présente contribution cherche à reconceptualiser certains enjeux fondamentaux du machine-learning en problématisant la question du « comportement » à l'aune de la théorie du cycle de l'image de Gilbert Simondon. En prenant au sérieux les positions théoriques et les ambitions techniques dans le domaine des algorithmes auto-apprenants, nous proposons de rendre compte de certaines transformations de l'algorithmique en la qualifiant de deux idéaux-types : mécanique et comportementale. Sans suggérer une quelconque opposition binaire, nous proposons plutôt de voir ces deux pôles comme des structurations successives, des phases d'existence qui transforment la nature du problème à résoudre, qui opèrent un passage de structure en structure. À la suite de Simondon, ce passage est requalifié de transductif , de façon à esquisser une image du comportement appelant une réinvention de nos cadres politiques et éthiques. Et si la question était : quel est le collectif qui advient, et que nous désirons voir advenir, quand nous nous comportons avec des algorithmes ?

Als Alternative zur scheinbaren Objektivität und Selbst-Verständlichkeit von ‹Daten› entwickelt vorliegender Aufsatz ein kritisches Konzept ‹digitaler Spuren›. Die Betonung der Materialität und Diskursivität von Spuren macht die technischen und soziopolitischen Implikationen digitaler Technologie verstehbar und aussagbar. Mit Gilbert Simondon und Michel Foucault werden die ontologischen und epistemologischen Rahmen geschaffen, um die Beziehungen zwischen Technologie und Prozessen der Subjektivierung deuten zu können. Digitale Spuren werden als Objekte und Produkte heteronomer Interventionen modelliert, deren Logik entlang der eingesetzten Algorithmen nachverfolgt werden kann. Das vorgeschlagene Paradigma eröffnet eine neue Perspektive darauf, wie Formen der Selbstkontrolle und der Kontrolle des Selbst voneinander abhängige Facetten einer ‹algorithmischen Gouvernementalität› sind. ; As an alternative to the seemingly natural objectivity and self-evidence of ‹data›, this paper develops a critical conceptualization of ‹digital traces›. Underlining the materiality and discursiveness of traces allows us to understand and articulate both the technical and sociopolitical implications of digital technology. The philosophies of Gilbert Simondon and Michel Foucault provide strong ontological and epistemological frameworks for interpreting the relationships between technology and processes of subjectification. In this light, digital traces are modeled as objects and products of heteronomous interventions, the logics of which can be traced through the algorithms deployed. The digital trace paradigm offers a new perspective on how forms of self-control and control of the self are interdependent facets of ‹algorithmic governmentality›.

This contribution aims at proposing a framework for articulating different kinds of "normativities" that are and can be attributed to "algorithmic systems." The technical normativity manifests itself through the lineage of technical objects. The norm expresses a technical scheme's becoming as it mutates through, but also resists, inventions. The genealogy of neural networks shall provide a powerful illustration of this dynamic by engaging with their concrete functioning as well as their unsuspected potentialities. The socio-technical normativity accounts for the manners in which engineers, as actors folded into socio-technical networks, willingly or unwittingly, infuse technical objects with values materialized in the system. Surveillance systems' design will serve here to instantiate the ongoing mediation through which algorithmic systems are endowed with specific capacities. The behavioral normativity is the normative activity, in which both organic and mechanical behaviors are actively participating, undoing the identification of machines with "norm following," and organisms with "norminstitution". This proposition productively accounts for the singularity of machine learning algorithms, explored here through the case of recommender systems. The paper will provide substantial discussions of the notions of "normative" by cutting across history and philosophy of science, legal, and critical theory, as well as "algorithmics," and by confronting our studies led in engineering laboratories with critical algorithm studies.