In his 2008 article, “The End of Theory: The Data Deluge Makes the Scientific Method Obsolete”, Wired editor Chris Anderson notoriously argues that the era of the scientific method built around testable hypotheses is fast-becoming obsolete due to the advent of “petabyte-scale” data accumulation and analytics. Observing that tech companies like Google treat this “massive corpus as a laboratory of the human condition” (Anderson 2008), Anderson polemically implores us to forget every theory of human behaviour, from linguistics and sociology to ontology and psychology. “With enough data”, he avers, “the numbers speak for themselves” (ibid.), and statistical correlation renders superfluous any theorisation of causality. The Californian Ideology (Barbrook and Cameron 1996)1 implicit in this overstatement of the benefits of data-driven research has been challenged by several commentators (see, e.g., Pigliucci 2009 and Mazzocchi 2015) and is patently specious. The appeal made by Silicon Valley zealots to ‘pay no attention to the man behind the curtain’ relies on a mythical, and sometimes even eschatologically feverish, conception of science as a purely objective pursuit, as well as a facile conception of human thought and behaviour—indeed, subjectivity—wherein we can be comprehensively understood via the purely quantitative approaches of Big Data analyses and the statistical aggregation of deep learning systems. As Louise Amoore underscores in Cloud Ethics, this ideology confronts our “fallible, intractable, fraught political world with a curious kind of infallibility. In the cloud, the promise is that everything can be rendered tractable, all political difficulty and uncertainty nonetheless actionable” (Amoore 2020, 55). This dream of complete mathematical and technological control over reality—of “a kind of atlas of clouds for the ineffable, a condensed trace of the trajectories of our future lives with one another” (ibid.)—is hardly new. What legal theorist Antoinette Rouvroy refers to as “algorithmic governmentality”2 (Rouvroy and Stiegler 2016, 6), by which she means “the increasingly statistical governance of the ‘real’ ensuing from a convergence of contemporary technological and socio-political evolutions” (Rouvroy 2011, 119), can be traced at least as far back as first-order cybernetics (Wiener 1965, 12) and, before that, to Leibniz’s calculus ratiocinator (cf. Couturat 1901). What makes the contemporary problem novel, however, is that the latest technologies, for example cloud computing, practically unlimited data storage, high-speed global communications networks and, most importantly, machine learning—and here we mean specifically new forms of connectionist ‘artificial intelligence’ that, unlike older symbolic AI models, rely on multi-layer artificial neural networks that are assumed to represent the biology of human cognitive structure—have powerfully exacerbated the quest for (and assumptions of) ‘Algorithmic Supremacy’. As digital technology theorist Dan McQuillan argues, data-driven modelling via statistical induction is assumed to bear inherent significance, but this approach, which McQuillan describes as a form of “machinic Neoplatonism” (McQuillan 2018), tends to entirely elide the broader subjective and inter-subjective contexts within which analyses and modelling unfold and is thus, via this unwarranted delimitation, able to present itself as operating with a level of mathematical objectivity it simply does not possess.
This creates several insidious problems. For one, as Bernard Steigler observes in The Age of Disruption, this deferral to the ‘superiority’ of algorithmic reason legitimises “the systematic exploitation and physical reticulation of interindividual and transindividual relations” in the service of the data economy (Stiegler 2019, 7). In other words, the kinds of transgenerational, intergenerational, interpersonal, personal and even pre-personal3 circuits that used to “emerge through affective relations of various kinds” over time within and across societies to forge “dreams, goals, objectives and common horizons” have been disrupted in unprecedented ways (ibid., 16). Rouvroy argues, in fact, that subjectivity is bypassed by contemporary digital automisation so that subjects are rendered little more than a “collection of infraindividual data” that are “recomposed at a supra-individual level under the form of profile” (Rouvroy and Stiegler 2016, 12). Second, the speed at which what we have elsewhere termed the Algocene4 has emerged, as well as the level of technological literacy a coherent understanding of this shift entails, has far exceeded our capacity to theorise its effects, with the result that many analyses of our contemporary condition remain trapped in anachronism, projecting the image of an old world and its struggles onto a new and largely alien terrain. It is perhaps not too hyperbolic to claim, as many have, that the staggering technological shifts of the last 20-odd years necessitate theoretical engagement on the scale of a new geological era or epoch. This is no longer the world of industrial capitalism, nor even of Foucault’s disciplinary societies (Foucault 1977), Deleuze’s control societies (Deleuze 1992) or Guattari’s Integrated World Capitalism (Guattari 2000), although these are all vestigially inflected in the new situation, which has variously, although far from exhaustively, been termed societies of hyper-control and computational capitalism (Stiegler 2019), cognitive capitalism (Yann Moulier-Boutang 2012), platform capitalism (Srnicek 2016), the age of planetary computerisation (Guattari 2013), The Stack (Bratton 2015), infopolitics (Koopman 2018), instrumentarianism (Zuboff 2019) and algorithmic governmentality (Rouvroy and Berns 2013). Third, because of the speed at which digital innovation operates, whatever effects this new “hyper-synchronization of consciousness” (Vignola 2017, 188) is having on subjectivity can only be perceived in the most miniscule of ways, yet we have to rely on this minimum of information for a symptomatology because, if we follow Deleuze via Nietzsche, it is only through a symptomatology that we can trace etiological factors in order to find the most inventive corrective therapy. A difficult task indeed considering the constraints we are outlining here! The point is, we urgently require new tools to grapple with the implications of the encroachment of algorithmic reason and governmentality into more and more aspects of our lives and minds. Far from eschewing theory in favour of the transcendent mathesis universalis of our new digital overlords, situated thinking about our times has never been more crucial. If, however, this kind of thinking has not been forthcoming to the extent we would hope for, then this too perhaps forms part of the symptomatology of the Algocene. In describing something that is unprecedented, we need to take care, as McKenzie Wark argues, to find “a renewed language for describing the present situation and identifying what in the received language of capitalism [or, for those further behind the times, Marxism] impedes forward movement in thought and action” (Wark 2019, 6). Having said this, we also hold that it is important not to reduce the entire social field to algorithmic control, despite its near-ubiquity. What is rather vital is to understand what this new way of experiencing subjectivity produces and what, in turn, produces it. As Guattari explains of the Freudian Unconscious, it produced new desires, which included “hysteria, infantile neurosis, psychosis, family conflict, the reading of myths, etc.” (Guattari 1995, 10). So too what Stiegler refers to as “negative collective protention”, or the nihilistic expectation of nothing, except perhaps The End (Stiegler 2019, 19, 50), has produced new desires, and here we have to seriously consider that The End might in fact—probably is—desired, even collectively so (especially given that we also face Anthropogenic doom).5