Affect Trapped: Algorithms, Control, Biopolitical Security

LIVING AND THINKING IN THE POSTDIGITAL WORLD Living and Thinking in the Postdigital World is the result of a series of conferences organized at the Collegium Artes Liberales, University of Warsaw, as a part of the project “Technology and Socialization”. Its main aim is to interrogate the different ways in which technology – especially digital technology – shapes today’s social and political landscape in a theoretical and practical way. The book is divided into three parts. The first one concentrates on theoretical elaborations of our current situation – testing if theories of technology that we have inherited from earlier ages are suited to our current historical moment. The second part of the book is devoted to describing novel experiences allowed by digital technologies and the intertwinement between our “online” and “offline” lives. The chapters gathered in the final part endeavor a look into the future, problematizing the consequences of currently observable trends and trying to understand the workings behind various visions of what is to come. LIVING AND THINKING IN THE POSTDIGITAL WORLD THEORIES EXPERIENCES EXPLORATIONS Edited by Szymon Wróbel Krzysztof Skonieczny Książka dostępna także jako e-book www.universitas.com.pl universitas LIVING AND THINKING IN THE POSTDIGITAL WORLD LIVING AND THINKING IN THE POSTDIGITAL WORLD THEORIES EXPERIENCES EXPLORATIONS Edited by Szymon Wróbel Krzysztof Skonieczny Kraków Publikacja dofinansowana przez Wydział„Artes Liberales” Uniwersytetu Warszawskiego Publication co-financed by the Faculty of „Artes Liberales”, University of Warsaw © Copyright for this edition by Towarzystwo Autorów i Wydawców Prac Naukowych UNIVERSITAS & University of Warsaw – Faculty of „Artes Liberales”, Kraków 2021 ISBN 978-83-242-3690-9 e-ISBN 978-83-242-6537-4 TAiWPN UNIVERSITAS Reviews Marta Bucholc, Assistant Professor, University of Warsaw Aleksandra Derra, Professor of Nicolaus Copernicus University in Toruń Proofreading Anna Olechowski Editing Magdalena Kusak Typesetting Stanisław Tuchołka / panbook.pl Cover design Sepielak Photography Michael Gaida, Pixabay www.universitas.com.pl Contents Szymon Wróbel, Krzysztof Skonieczny, Introduction . . . . . . . . . . . 7 Part One. Theorizing the Technological Present . . . . . . . . . . . . . . . 13 Szymon Wróbel, Dismantling the Concept of Technology . . . . . . . . . . 15 Ivan Dimitrijević, Judgement upon Work . . . . . . . . . . . . . . . . . . . . . . 35 Michael Stemerowicz, “The Work of Art in the Age of Its Technological Reproducibility” – A Rehearsal of Consumerism and the Aesthetic Consequences of the Dissolution of Tradition . . . . . . . . . . . . . . . . . . 57 Mümtaz Murat Kök, Optimism as Attachment to Capitalism . . . . . . 77 Adam Lipszyc, Affect Unchained: Violence, Voyeurism, and Affection in the Art of Quentin Tarantino . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Part Two. Experiencing the Technological Present . . . . . . . . . . . . 109 Michaela Fišerová, Touching and Retouching. Question of Authenticity in Social Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Agata Szepe, Israeli Tourism to Poland in Social Media. Perspectives of Social Science of the Internet and the Actor-Network Theory Approach . . . . . . . . . . . . . . . . . . . . . 127 James W. Besse, Designing Emotional Styles . . . . . . . . . . . . . . . . . . . 141 Joanna Łapińska, Posthuman and Post-Cinematic Affect in ASMR “Fixing You” Videos . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 6 Contents Julia Krzesicka, Emotional Capitalism, Sociability, and Orality. Speculative Imagination Exercise on the Future of Work for Voice Assistants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Karin-Ulrike Nennstiel, A Social Robot in Your Living(room)? Recent Developments and Shifting Appraisals . . . . . . . . . . . . . . . . . 185 Part Three. Exploring Technological Futures . . . . . . . . . . . . . . . . . 203 Krzysztof Skonieczny, Deleuze’s Remarks on Control Societies. Consequences for Work and Education . . . . . . . . . . . . . . . . . . . . . . 205 Denis Petrina, Affect Trapped: Algorithms, Control, Biopolitical Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Ewa Mazierska, Representation of Postdigital Encounters in Recent Science Fiction Films . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Adam Cichoń, Dr. Strange(love) or: From Affection-Images to Inter-Faces . . . . . . . . . . . . . . . . . . . . . . . 253 Mitchell Atkinson III, Substrate Independence, Migration, and the Naturalistic Attitude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 About the authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Denis Petrina Affect Trapped: Algorithms, Control, Biopolitical Security ABSTRACT In 1992, Gilles Deleuze established a frame to analyze new taxonomies of power, based less on disciplinary power than on “free-floating” control. Technologically, this novel society of control breaks from disciplinary societies that mainly relied on energy-producing mechanisms, employing much more complex machines – computers. This paper argues that the shift from machines to computers is not merely a technologically induced reconfiguration of the socioeconomic field but, more importantly, a biopolitical one, for it establishes a new regime that targets, governs, and controls subjects, their minds and bodies – that is, their affects. In the digital environment, the latter is captured and transformed by algorithms – the operational principle of the machines mentioned above. Thus, this paper aims to examine a biopolitical framing of algorithm-driven media closely, discuss how algorithms function as biopower mechanisms, and, finally, raise a question whether (and how) it would be possible to resist becoming algorithmic biopower. KEYWORDS: algorithms, affect, control, apparatus of security, biopolitics Introduction: From Algorithms to Power Recently, “algorithm” has become a buzzword. From scientific disciplines (cybernetics, medicine, economics) to business environments, today algorithms have become utterly omnipresent as a procedure (for lack of a better term), integrated into systems, for data extraction and processing, calculation and analysis, the indication of problems and their solutions, and a broad spectrum of other functions. Bernard Chazelle, a prominent computer scientist, has even termed algorithms “the idiom of modern science”, placing particular importance on understanding how algorithms work, as sciences do 220 Denis Petrina and perhaps will change, while, Chazelle believes, algorithms are to stay (Chazelle 2006). The innovative impetus of algorithms, as their numerous applications showcase, lies, quite simply, in their independence from the human operator (which has been translated into terms of [artificial] intelligence), and thus in their full automation, high performance, and adaptability. More importantly, algorithms do what human beings cannot do or do very poorly: they are programmed and are, therefore, fully capable of operating in the conditions of uncertainty, estimating likelihood and making “intelligent” decisions based on the input data they receive. Attempting to theorize the features of algorithms and place them in a social context, the automated (even autonomous?) nature of algorithms signifies a conclusive break from the so-called “Fordist socioeconomic model”. It symbolizes, if not embodies, the post-Fordist logic of (economic, social, and cultural) production. While the former is conventionally associated with standardization, mass production, and consumption – in short, homogeneity – the latter is typically described as far less standardized and more flexible, characterized by economies of scope and, as a result, as inherently heterogeneous and dynamic. We might even dare to say that this was brought up by the algorithmization of radical reorientation from the rigid Fordist culture towards less regulated and more diversified post-Fordist environment (both material and digital) that allows (or at least promises) us to escape from the disciplinary apparatus of work, restrictive society, and even the dominating model of thinking. But does it, actually? Epistemologically and practically, as discussed above, algorithms have, indeed, been heralded as a breakthrough in the way we think about things and do them. However, is this breakthrough as emancipatory (again, epistemologically, practically, and even politically) as apologists of algorithms tend to view it? While obviously restraining from technological pessimism, I tend to view algorithms, as well as their effect on contemporary culture, critically – and offer a more nuanced and problematic philosophical perspective to look at them. If we were to (re-)read Gilles Deleuze’s (1992) Postscript on the Societies of Control, we could not help but notice how, upon closer inspection, algorithms staggeringly resemble a new regime of power that Deleuze grasps conceptually. Also, if we were to pay closer attention Affect Trapped: Algorithms, Control, Biopolitical Security to how Deleuze frames the discussion of the evolution of taxonomies of power, we would be able to discern the crucial role of technologies in the development of the socioeconomics of control. Indeed, algorithms are not mechanic entities, they are not machines, and they are not hardware that is representative of the Fordist socioeconomy; instead, they represent the principle of operation – in other words, they are software. This, again, makes us rethink not only practical techne – how things are done; but also, most importantly, political techne – which tools power utilizes. Here, two conclusions are possible. The first one is that the shift from hardware – machines – towards software – principles – is emblematic of what, following Michel Foucault, could be called “vaporization of power”: power becoming less visible, less tangible, and, generally, less perturbing. This leads to the second, undoubtedly naïve, conclusion that vaporization of power is tantamount to the vanishing of power. Quite the contrary, Foucault’s lesson is that the vaporization of power means nothing but the fact that power in contemporary societies is everywhere – it is the very air we breathe (Foucault 1995, 26). As insightfully marked by Bruce Curtis, decentralization (qua already discussed “vaporization”) of sovereign power does not amount to the absolute disappearance of power – in fact, it is an extension of what Foucault terms “governmentality”, a macabre combination of government and rationality (Curtis 2002, 506). This remark leads us to another critically important Foucault’s notion – “biopolitics”. For Foucault, the biopolitical regime (which he links with neoliberal rationality) is not governmental in its traditional sense, but rather environmental: it is not the disciplinary regime that seeks to target individuals, drill and punish them, but a much more complex ecology of powers (Foucault 2003, 246). This paper attempts to argue that algorithms and, consequently, algorithmic rationality are, indeed, an integral part of this new ecology of powers marking the emergence of the societies of control. Thus, it will offer a closer examination of the latter and will argue that control is algorithm-driven per se. It will also reveal the links between the biopolitical apparatus of security and algorithmic governmentality/environmentality whose target, as it will be explicated further, appears to be an “affect”, cognitive and physical bodily capacities. Ultimately, potential scenarios of resistance (affect-invested and affect-driven) will be discussed. 221 222 Denis Petrina From Discipline to Control Following Patricia Clough, the break between the ancien régime and nouveau régime could be conceptualized as a transition from discipline and representation towards control and information (Clough 2007, 14). Deleuze’s almost prophetic text, written in 1990, Postscript on the Societies of Control, outlines the genealogical development of control, tracing it back to disciplinary societies, which, in their turn, developed from sovereign societies. In his essay, Deleuze develops a detailed analytical framework for outlining essential differences between two sociopolitical regimes, focusing on such aspects as (1) spatial environments, (2) socioeconomic models, (3) dynamics, (4) subjectivity, (5) scientific implications, (6) machines, all of which are discussed further. As Postscript on the Societies of Control attempts to expand Foucault’s critical historiography (genealogy) and at the same time digresses from it, the essay is deeply informed by Foucauldian theory, borrows vocabulary from it, interprets, and reinterprets it. For instance, when discussing disciplinary societies, Deleuze associates them with “environments of enclosure”, such as a prison, hospital, factory, school, and family (Deleuze 1992, 4). This topology of enclosure, as Deleuze explains it and as Foucault elaborates in his writings, is tightly linked with institutions, in which an individual performs a certain role (a student in school, a worker in a factory, etc.), and could, therefore, be characterized by breaks and interruptions – in short, stratification. The latter, however, should be construed as a “repertoire” of identities: again, the student, the worker, the prisoner, which are more or less similar for all individuals placed in the same context. That is why Deleuze argues that the socioeconomic model of Fordist disciplinary societies is “a factory” (Deleuze 1992, 5), which, according to Foucault, presupposes a Panoptic logic of close supervision and tireless control exercised upon agents, each of whom is an integral part of a broader mechanism and performs their functions in accordance with strict regulations and norms (Foucault 1995, 174-175). Seen in this light, disciplinary societies rely heavily on the subjectivity; yet, again, not single individuals, but rather, as Deleuze explains, “an individual within a mass” (Deleuze 1992, 5). Even though individual behavior is supervised, tracked, corrected, and controlled, Affect Trapped: Algorithms, Control, Biopolitical Security it matters only as a means that makes extrapolation possible: that is, it allows to establish a norm, which becomes a correlative of productivity in economy and discipline in society. Since the nexus between productivity and discipline is incorporated in almost every sphere of life in disciplinary societies and is inseparable from them, the epistemological framework of these could be named as “mechanics”, and more precisely, “thermodynamics”. Put simply, the object of these disciplines is the bodies and energy they produce; discipline is, indeed, concerned with the production of (Foucault would call them “docile”) bodies as well as the reproduction of them, while productivity directly correlates with extraction, conversion, and distribution of energy among different parts to make the whole productive. Thus, these elements, all together, constitute “capitalism of production” (Deleuze 1992, 6). In societies of control, a Fordist model of “capitalism of production” is substituted for “capitalism of marketing” (the utmost form of commodity fetishism, in which commodities become more “alive” than humans), which Deleuze himself calls “the soul of the corporation” (Deleuze 1992, 6). The corporation with a soul is the new socioeconomic model that diverges from the “heartless” factory: unlike the latter, the former does not depend upon institutions (in fact, it signifies the end of institutions), and instead of relying on various environments of enclosure, creates and fosters an open environment, where, as Deleuze notices, control is “free-floating” (Deleuze 1992, 6). It is therefore characterized by perpetuality and continuousness, which contribute to its reflexivity – as Scott Lash explains, this metaphysical (rather than its predecessor, physical) form of capitalism is best represented by cybernetic systems that are reflexive and self-organizing due to their openness to the environment (Lash 2007, 8). For this reason, physics, mechanics, and thermodynamics are all insufficient epistemological frameworks for understanding control societies, in which corporations have souls and capitalism is metaphysical. Extracting and transmitting information as well as developing and regulating networks are two practices that constitute the backbone of control societies, which makes cybernetics their prototypical, epistemological paradigm. Not only does the cybernetic nature of control societies weaken institutional power, but it also discards the product of institutions – 223 224 Denis Petrina individuals and masses as a form of subjectivity. The question of subjectivity is highly problematic in the context of metaphysical/cybernetic/control-based capitalism: who is the subject if all that we know about them is the scattered segments of their data? Deleuze skillfully answers this question by proclaiming the notion of the “individual” as obsolete (or, to be more precise, attributing it solely to disciplinary societies) and insisting that we have become “dividuals” instead – a collective concept that encompasses “masses, samples, data, markets, or banks” (Deleuze 1992, 5) – in short, data. To reiterate this, the data is the subject (and vice versa); this statement radically reshapes the philosophical understanding of the subject, traditionally understood as a coherent and unified entity, and proposes the alternative concept of the subject, which, following John Cheney-Lippold, might be described as an “algorithmic identity, [which is] an identity formation that works through mathematical algorithms to infer categories of identity on otherwise anonymous being” (Cheney-Lippold 2011, 165). From Control to Computers (and Back) As demonstrated, one of the core aspects of the Deleuzian philosophical methodology is what might be termed “diagramization” of social relations and their underlying principles. The diagram, however, should not be conceived as a visual representation of a certain aspect of the complexity of the social assemblage (for instance, “the factory” is not the representation of the actual manufacturing facility), but rather, as Manuel De Landa helpfully explains, as a complex system that accounts for virtual singularities (De Landa 2000, 33). Alexander Galloway maintains: just as the fold was Deleuze’s diagram for the modern subject of the baroque period, the superfold is the new “active mechanism” for life within computerized control society. (Galloway 2012, 524) And even more so: to understand the control society we live in, it is imperative to recognize the “computer as its central mitigating factor” (Galloway 2012, 524). In other words, the computer – qua the superfold – is the diagram of control societies. Affect Trapped: Algorithms, Control, Biopolitical Security Its genealogic predecessor is any type of machine that involves energy (again, the focus is placed on productivity and thermodynamic), which Deleuze associates with disciplinary societies. The danger for these machines takes the form of entropy, which potentially disrupts the order of disciplinary societies (Deleuze 1992, 6). Interestingly enough, in the machinic diagram of control societies, the computer operates in the environment of entropy, cybernetic chaos, and abundance of structureless information. Unlike energy-driven machines, which are closed and in this regard rather simple in terms of their structure, computers are complex: their complexity comes from the order within which they operate; while capitalism of production qua Fordist capitalism qua disciplinary societies rely on easily identifiable entities: materials, property, products, computerized societies of control produce and reproduce “coded figures” that are inherently “deformable and transformable” (Deleuze 1992, 6). In short, computers are the very emblem of the shift from the danger of entropy towards systematic, productive, and the alarming exploitation of information disorder. We might, however, ask a reasonable question: what is the relation between computers and control? Perhaps to answer it, we need to determine what Deleuze actually means by the illusive term “control”. Galloway reminds us that English and French meanings of the word “control” differ: while being rather definite in English, in French, control presupposes certain “monitoring apparatuses, such as train turnstiles, border crossings and check points” (Galloway 2012, 522), which, in their turn, let us associate control with such characteristics, as mobility, openness, and (controlled) liberation. In this sense, control is not as much political as it is biopolitical – and, as Clough argues, digital media and computational technologies are, indeed, deeply embedded in biopolitics (Clough 2018, xviii). From Computers to Apparatuses of Security (Algorithms) Foucault’s account of biopolitics mirrors many aspects of control societies, thoroughly examined by Deleuze. Perhaps it would be safe to say that Foucauldian biopolitics is control-oriented as much as control societies are biopolitical. The first reason to consider them together 225 226 Denis Petrina is that both the biopolitical regime and societies of control emerge as a response to a problem of entropy. Eugene Thacker notes that biopolitics as a mode of governmentality appeared when governments became preoccupied with the question of control – more precisely, the management and regulation of multiplicities; “and the problem is greater when the multiplicities in question are construed as living multiplicities”, he writes (Thacker 2011, 152; the emphasis is his). Thus, the object of power undergoes radical transformation – if disciplinary regimes attempted to govern individual bodies, biopolitics targets whole populations (“living multiplicities”). Understandably, these multiplicities could be encountered by governmental apparatuses only through a means of representation: in case of populations, statistical representation – which, translated into the vocabulary of societies of control, is nothing but raw data. This data represents “not the actual totality of the subjects in every single detail, but the population with its specific phenomena and processes” (Foucault 2009, 93). Obviously, it is impossible to govern or discipline “phenomena and processes”, yet it is possible to control them: and here, control is synonymous with security, or, to be more precise, the apparatus of security. Governmentality – the art of governing – is of utmost importance here: for security and control tend to mask themselves, they appear as a modulation of laissez-faire or not governing too much (Foucault 2008, 102) – and should thus modify the very environment they operate in to be efficient. This makes security and control not only governmental but, even more importantly, environmental, since, as Brian Massumi explains, it regulates causes rather than effects (Massumi 2015, 22). Therefore, the old principle of normation (exemplary of disciplinary societies) does not apply here; the proper term here seems to be “normalization” (Foucault 2008, 146). The difference lies in the strategy how techniques and technologies of power are applied: in disciplinary societies, problems (delinquency, poverty, epidemics) were solved by correcting problematic individuals/bodies; in biopolitical societies of control, however, eradication of problems is not an end in itself: instead, the limit is set that makes the problem tolerable. “Optimality” is another word for the principle of laissez-faire: its conceptual formula is “neither too much nor not enough” – just the right amount. The logic of normalization is algorithmic per se, Affect Trapped: Algorithms, Control, Biopolitical Security as it is concerned with statistically-derived mathematical optimization (the selection of the most representative and relevant elements from sets of data), risk assessment and mitigation (not elimination though), and, ultimately, constant re-evaluation, semi-automated learning, and modeling. According to Maurizio Lazzarato, the contemporary regime is comprised of three major trajectories of governing, each of which targets a specific domain of social and personal life. This triad includes three “Ms”: management of population (biopower) through the molding of the individual (discipline) and modulation of mind (control) (Lazzarato 2006, 176-177). Algorithms are their point of intersection: as discussed above, biopolitical management of populations is algorithmic in nature, as it heavily relies on representational data and is constantly transformed by it; discipline allows the apparatus of security to extract the necessary data from bodies and individuals and in so doing “fuels” the apparatus; and, finally, control is technically enabled by computers where data is circulated. The digital data becomes “the very ‘texture’ of capitalism” and results in the mutation of capitalism, already discussed by Deleuze: capitalism of production transforms into what Antoinette Rouvroy terms “algorithmic governmentality” (Rouvroy 2016, 30-31). Such governmentality largely exploits what, after Clough, might be called the “user unconscious”: a strange blend of subjective unconscious mechanisms commonly shared by users of a specific medium which produce “technical substrates” (Clough 2018, x). Put simply, users are unaware of the functional mechanism behind the technical facilities they use, which in algorithm studies has been conceptualized as the problem of “the black box”: the problem of opaqueness, which prevents us from knowing why an input X produces an output Y. Viewed through this problematic lens, the algorithmic principle of operation of biopower could also mean that algorithms are potentially biopowerful. If this is the case, it begs the question: how precisely are algorithms utilized as exploitative tools of (bio)power? From Algorithms to Affects Tarleton Gillespie, a principal researcher at Microsoft, defines algorithms as follows: 227 228 Denis Petrina they are encoded procedures for transforming input data into a desired output, based on specified calculations. The procedures name both a problem and the steps by which it should be solved. (Gillespie 2014, 167) Semantically, this definition omits a crucial component for a fuller understanding of how algorithms operate: the agent. Who is the agent encoding procedures? More importantly, what is a “desired output” (desired by whom?) and who decides what the “problem” that has to be solved is? Even though the clarity of the description of procedures achieved by the use of the passive voice signifies that algorithms are technologically neutral (that is, this technology is not “evil” per se), this description suggests that basically anyone (provided they have the necessary knowledge and skills) can be the agent. And, taking into account the fact that biopolitics is algorithmic, it becomes apparent that algorithms are not politically neutral. The seme of “calculation” reveals that algorithms are perfectly suitable tools for biopower, since the latter strives, as it has been discussed, to set limits, establish optimal conditions, evaluate and prevent losses, optimize costs, etc. What is more, both mathematical algorithms and statistical biopower target the same object: data. However, what is data? When does a body, a living being, become data? What mediates technologically intelligible data – or, more precisely, big data, large sets produced by numerous populations data – and living multiplicities? The answer, however paradoxical it may seem, is affects – intertwined cognitive and bodily states, whose derivatives are feelings, moods, emotions on the one hand and capacities, actions, and interactions on the other. Clough explains that “under datalogical conditions, measurement is always a singularity – a productive, affective materialization of dynamics and relations of recombinable forces” (Clough 2018, 108). And yet, even though the singularity of an affect – a particular action/reaction of a particular individual – is treated as a singularity, it is then subsumed under broader categories representative of the whole population, which, indeed, makes it algorithmic. To clarify, algorithms operate under the conditions of uncertainty. Affects are, indeed, characterized by uncertainty: there is no way to figure out how a body would react to a certain trigger and what the outcome would be. Or is there? If the modus operandi Affect Trapped: Algorithms, Control, Biopolitical Security of biopolitical control is the triad of risk evaluation, limitation, and optimization, and if algorithms are procedures that are capable of “transforming input data [no matter how ‘uncertain’ it is] into a desired output” (Gillespie 2014, 167), this task does not seem unsolvable whatsoever. If the body (mediated through affect and then through data) does not react to a trigger in a desired manner, it can be trained to react so. It is important to understand, however, that algorithms are not magical “hypodermic needles”: they are not “injected” directly into the bodies or brains to produce “the desired outcome”; obviously, the process is not as linear as we might imagine and much more complex. Biopower 2.0 operates in the middle space, as Cheney-Lippold, following Foucault, explains: this is a peculiar type of space, “where the modulation of categorical identity changes according to a ‘culture, imperceptibly deviating from the empirical orders prescribed for it by its primary codes’ (Foucault 1973, xx)” (Cheney-Lippold 2011, 174). This is the type of operation that Clough has revealed: a “black box” of recombination, in which “affective circulation” (Swenson 2011, 20) takes place: affects are captured, encoded as intelligible data, and then this data is used to produce a desired output – or desired affects. Ultimately, the very apparatus (algorithm) is changed: since, as it has been continuously mentioned, it is automated and adaptable, not only does it know (qua is programmed) how to capture needed affects and (en/de)code them, but it also learns from the coded data and transforms itself to better function in the conditions of uncertainty. Gillespie, too, has emphasized that algorithms can be trained to “identify qualities within the data that might help it to discern different types of content” (Gillespie 2018, 216). Different types of content are then categorized; Cheney-Lippold views such categorization as a creation of digital “ontologies (…) embedded within a set of power relations” (Cheney-Lippold 2011, 174). Further, he argues that this new form of biopower regulates how these digitally constructed quasi-ontological categories define life (Cheney-Lippold 2011, 175) as a continuum of affective actions and interactions. Viewed broadly, affective investments (governed by algorithms) become affective re-investments; for example, collected from users marketing-related data exposes hedonic aspects of human life, such as interests, tastes, even desires, which are then shaped 229 230 Denis Petrina and reshaped on the basis of the interaction with the algorithms. Purchase data may be used to determine a user’s economic status, as well as directly influence their decisions to buy (again, thanks to the algorithmically constructed “category” for the population this user potentially represents; if a person does not make a purchase, algorithms will figure out the reason of such a decision and will therefore adjust). Even more so, algorithms, in a very worrying biopolitical manner, have direct access to individuals’ bodies: all types of wearables, heart rate trackers, and similar types of smart technology process myriads of bytes of information derived directly from a human body – from the processes we ourselves may not be fully aware of. Thus, this new topology of control – the middle space CheneyLippold speaks of – emerges at the intersection of biology, politics, and technology. Indisputably, it calls for the reconsideration of the category of the body and its affects. A concept of biomedia, coined and developed by Thacker, is of particular help here: according to Thacker, the informatic capacity to affect biological materiality (the very blend of cybernetic and biological) results in the recontextualization of a body that becomes “more than a body”, since “the body you get back is not the body with which you began” (Thacker 2003, 53). The space – the “middle space” where transformations of the body occur – where its affects are extracted and then re/invested – is the opaque “black box”. This, on the one hand, allows us to view the biomediated body as experimentally transcendental (in a sense that the limits of the body are exceeded far beyond its physioanatomical functions), but, on the other, it appears to be extremely susceptible to biopower. Conclusion: From Affects to Knowledge The problem of “black box” prompts another, more serious problem – the problem of epistemic injustice, hindering our capacities “as a subject of knowledge” (Fricker 2007, 5). The best way to express this problem is rather simple: algorithms know about us more than we know about them. Obviously, algorithmic rationality has drastically transformed the cultural, social, political landscape we inhabit today, as well as the very way we think about and understand it. This Affect Trapped: Algorithms, Control, Biopolitical Security leads Paško Bilić to believe that particular emphasis should be placed on algorithmic literacy, accompanied by transparency and oversight (Bilić 2018, 327). This is, however, not a case of “know thy enemy”, but rather an invitation to acknowledge the change, interpret it in a variety of ways, find means to comprehend it – and, most importantly, take active part in it. After all, the logic of this change, as it has been noted, is affective, and affects are not only the change in our bodies we experience but also, create. However anachronistic it may sound, in conclusion, I would like to refer to the most prominent theorist of affects of the 17th century (if not the most prominent of all times) – Baruch Spinoza. As epistemic injustice – unequal distribution of knowledge among individuals – is directly linked to freedom, Spinoza’s remarks on the relation between knowledge and freedom seem particularly insightful. In the note to the tenth proposition of the fifth part of his Ethics, he states that knowledge is liberating as long as we learn how certain things affect us – and, vice versa, how we affect others (Spinoza 2001, 115). Such a type of active, praxis-oriented epistemology, whose ultimate goal is liberated knowledge, teaches us that we are an integral part of the world we live in and thus fosters curiosity and openness to the world, as well as encourages us to take an active part in building this world, being responsible for ourselves and others. The ultimate lesson we can learn from Spinoza is that knowledge, freedom, responsibility, and joy are four corners of the same square. In a more contemporary context, recent attempts to escape the claws of biopolitical capitalism – the advent of systems of decentralized control, the principle of net neutrality, socially responsible hacktivism – all herald a productive reconciliation of contemporary technology and emancipatory aims. All these and many more examples have been inspired by a totally different social imaginary from the one that is constructed, both physically and digitally, by a capitalist regime; all these examples represent the urge to bring beyond speculative imagination societies that, from the ruins of dismantled hierarchies, would emerge as horizontal social spaces, societies that would be based on sharing instead of owning, and on mutual responsibility and commonality instead of isolation and dividuality. The true takeaway question is whether algorithms (as they are now) would potentially have a place in this idyllic infrastructure. 231 232 Denis Petrina References Bilić, Paško. 2018. “A Critique of the Political Economy of Algorithms: A Brief History of Google’s Technological Rationality.” TripleC 16 (1): 315-331. Chazelle, Bernard. 2006. “The Algorithm: Idiom of Modern Science.” Accessed January 14, 2020. https://www.cs.princeton.edu/~chazelle/ pubs/algorithm-print.pdf. Cheney-Lippold, John. 2011. “A New Algorithmic Identity: Soft Biopolitics and the Modulation of Control.” Theory, Culture and Society 28 (6): 164-181. Clough, Patricia. 2007. “Introduction.” In The Affective Turn: Theorizing the Social, edited by Patricia T. Clough, and Jean Halley, 1-33. Durham and London: Duke University Press. Clough, Patricia. 2018. The User Unconscious: On Affect, Media, and Measure. Minneapolis and London: University of Minnesota Press. Curtis, Bruce. 2002. “Foucault on Governmentality and Population: The Impossible Discovery.” The Canadian Journal of Sociology 27 (4): 505-533. De Landa, Manuel. 2000. “Deleuze, Diagrams, and the Genesis of Form.” American Studies 45 (1): 33-41. Deleuze, Gilles. 1992. “Postscript on the Societies of Control.” October 59: 3-7. Foucault, Michel. 1973. The Order of Things: An Archaeology of the Human Sciences. New York: Vintage Books. Foucault, Michel. 1995. Discipline and Punish: The Birth of the Prison. Translated by Alan Sheridan. New York: Vintage Books. Foucault, Michel. 2003. Society Must Be Defended: Lectures at the Collège de France 1975-1976. Translated by David Macey. New York: Picador. Foucault, Michel. 2008. The Birth of Biopolitics: Lectures at the Collège de France 1978-1979. Translated by Graham Burchell. New York: Palgrave Macmillan. Foucault, Michel. 2009. Security, Territory, Population: Lectures at the Collège de France 1977-1978. Translated by Graham Burchell. New York: Palgrave Macmillan. Fricker, Miranda. 2007. Epistemic Injustice: Power and the Ethics of Knowing. New York: Oxford University Press. Galloway, Alexander. 2012. “Computers and the Superfold.” Deleuze and Guatarri Studies 6 (4): 513-528. Gillespie, Tarleton. 2014. “The Relevance of Algorithms.” In Media Technologies: Essays on Communication, Materiality, and Society, edited by Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten A. Foot, 167-193. Cambridge and London: The MIT Press. Affect Trapped: Algorithms, Control, Biopolitical Security Gillespie, Tarleton. 2018. Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions that Shape Social Media. New Haven and London: Yale University Press. Lash, Scott. 2007. “Capitalism and Metaphysics.” Theory, Culture and Society 24 (5): 1-26. Lazzarato, Maurizio. 2006. “The Concepts of Life and the Living in the Societies of Control.” In Deleuze and the Social, edited by Martin Fugslang, and Bent Meier Sørensen, 171-190. Edinburgh: Edinburgh University Press. Massumi, Brian. 2015. Ontopower: War, Powers, and the State of Perception. Durham and London: Duke University Press. Rouvroy, Antoinette. 2016. “Algorithmic Governmentality: Radicalization and Immune Strategy of Capitalism and Neoliberalism?” La Deleuziana – Online Journal of Philosophy 3: 30-36. Spinoza, Baruch. 2001. The Ethics. Translated by Robert Harvey Monro Elwes. [Blakmask Online.] Swenson, Kristin. 2011. “Affective Labor and Governmental Policy: George W. Bush’s New Freedom Commission on Mental Health.” Baltic Journal of Law and Politics 4 (2): 1-23. Thacker, Eugene. 2003. “What is Biomedia?” Configurations 11 (1): 47-79. Thacker, Eugene. 2011. “Necrologies; or, the Death of the Body Politic.” In Beyond Biopolitics: Essays on the Governance of Life and Death, edited by Patricia T. Clough, Jean Halley, 139-162. Durham and London: Duke University Press. 233