Questo saggio è il risultato di un’attività di ricerca che coinvolge una serie di istituzioni italiane di autoformazione di ispirazione post-operaista, di studiosi e ricercatori di lingua inglese impegnati nell’analisi dei media sociali e della teoria dei media digitali e anche artisti, attivisti, lavoratori della conoscenza, precari et similia. Trae spunto dal seminario che si è svolto a Londra il 20 gennaio 2014, ospitato dall’Unità di Cultura Digitale presso il Centro per gli Studi Culturali (Goldsmiths ‘ College, University of London). Il workshop è stato il risultato di un processo di riflessione e di dibattito che è iniziato all’intrno della rete Uninomade 2.0 nei primi mesi del 2013 e proseguita attraverso mailing list e siti web come Euronomade ( http://www.euronomade.info/), questo sito di Effimera, Commonware (http://www.commonware.org/ ) , i Quaderni di San Precario (http://quaderni.sanprecario.info/) e altri. Più che un saggio tradizionale, quindi , si propone di essere un sintetico documento, che, nato all’interno di una ‘ rete di ricerca sociale’, ha lo scopo di discutere una serie di problemi, tesi e nodi tra teoria politica e ricerca nel campo della scienza, della tecnologia e delle nuove forme della valorizzazione capitalistica.(pubblicato in contemporanea su Euronomade)
* * * * *
Premise
This essay is the outcome of a research process which involves a series of Italian institutions of autoformazione of post-autonomist inspiration (‘free’ universities engaged in grassroots organization of public seminars, conferences, workshops etc) and anglophone social networks of scholars and researchers engaging with digital media theory and practice officially affiliated with universities, journals and research centres, but also artists, activists, precarious knowledge workers and such likes. It refers to a workshop which took place in London in January 2014, hosted by the Digital Culture Unit at the Centre for Cultural Studies (Goldsmiths’ College, University of London). The workshop was the outcome of a process of reflection and organization that started with the Italian free university collective Uninomade 2.0 in early 2013 and continued across mailing lists and websites such as Euronomade (http://www.euronomade.info/), Effemera, Commonware (http://www.commonware.org/), I quaderni di San Precario (http://quaderni.sanprecario.info/) and others. More than a traditional essay, then, it aims to be a synthetic but hopefully also inventive document which plunges into a distributed ‘social research network’ articulating a series of problems, theses and concerns at the crossing between political theory and research into science, technology and capitalism.
What is at stake, then, is the relationship between ‘algorithms and ‘capital’, that is the increasing centrality, announced in the document that called for the workshop, of algorithms ‘to organizational practices arising out of the centrality of information and communication technologies stretching all the way from production to circulation, from industrial logistics to financial speculation, from urban planning and design to social communication.’ (http://quaderni.sanprecario.info/2014/01/workshop-algorithms/). These apparently esoteric mathematical structures, have also become part of the daily life of users of contemporary digital and networked media. Most users of the Internet daily interface or are subjected to the powers of algorithms such as Google’s Pagerank (which sorts the results of our search queries) or Facebook Edgerank (which automatically decides in which order we should get our news on our feed) not to talk about the many other less known algorithms (Appinions, Klout, Hummingbird, PKC, Perlin noise, Cinematch, KDP Select and many more) which modulate our relationship with data, digital devices and each other. This widespread presence of algorithms in the daily life of digital culture, however, is only one of the expressions of the pervasiveness of computational techniques as they become increasingly co-extensive with processes of production, consumption and distribution displayed in logistics, finance, architecture, medicine, urban planning, infographics, advertising, dating, gaming, publishing and all kinds of creative expressions (music, graphics, dance etc).
The staging of the encounter between ‘algorithms’ and ‘capital’ as a political problem invokes the possibility of breaking with the spell of ‘capitalist realism’, that is the idea that capitalism constitutes the only possible economy while at the same time claiming that new ways of organizing the production and distribution of wealth need to seize on scientific and technological developments (Fisher 2009, Negri 2014b). Going beyond the opposition between state and market, public and private, the concept of the common is used here as a way to instigate the thought and practice of a possible post-capitalist mode of existence for networked digital media.
Algorithms, capital and automation
Looking at algorithms from a perspective that looks at the constitution of a new political rationality around the concept of the ‘common’ means engaging with the ways in which algorithms are deeply implicated with the changing nature of automation. Automation is described by Marx as a process of absorption in the machine of the ‘general productive forces of the social brain’ such as ‘knowledge and skills’ (Marx 1973: 694), which hence appear as an attribute of capital rather than the product of social labour. Looking at the history of the implication of capital and technology, it is clear how automation has evolved away from the thermo-mechanical model of the early industrial assembly line toward the electro-computational dispersed networks of contemporary capitalism. It is possible hence to read algorithms as part of a genealogical line that, as Marx put it in the ‘Fragment on Machines’ starting with the adoption of technology by capitalism as fixed capital, pushes the former through several metamorphoses ‘whose culmination is the machine, or rather, an automatic system of machinery.. set in motion by an automaton, a moving power that moves itself’ (Marx 1973: 692). The industrial automaton was clearly thermodynamical and gave rise to a system ‘consisting of numerous mechanical and intellectual organs so that workers themselves are cast merely as its conscious linkages'( ibidem). The digital automaton, however, is electro-computational, it puts ‘the soul at work’ and involves primarily the nervous system and the brain and comprising ‘possibilities of virtuality, simulation, abstraction, feedback and autonomous processes’ (Fuller 2008: 4; Berardi). The digital automaton unfolds in networks consisting of electronic and nervous connections so that users themselves are cast as quasi-automatic relays of a ceaseless information flow. It is in this wider assemblage, then, that algorithms need to be located when discussing the new modes of automation.
Quoting a textbook of computer science, Andrew Goffey describes algorithms as ‘the unifying concept for all the activities which computer scientists engage in… and the fundamental entity with which computer scientists operate.” (Goffey 2008: 15) An algorithm can be provisionally defined as the “description of the method by which a task is to be accomplished…’ by means of sequences of steps or instructions, sets of ordered steps that operate on data and computational structures. As such, an algorithm is an abstraction, ‘having an autonomous existence independent of what computer scientists like to refer to as “implementation details,” that is, its embodiment in a particular programming language for a particular machine architecture’ (Goffey 2008: 15). It can stretch in complexity from the most simple set of rules described in natural languages (such as those used to generate coordinated patterns of movement in smart mobs) to the most complex mathematical formulas involving all kind of variables (as in the famous Monte Carlo algorithm used to solve problem in nuclear physics and later also applied to stock markets and now to study non-linear technological diffusion processes). At the same time, in order to work, algorithms must exist as part of assemblages that include hardware, data, data structures (such as lists, databases, memory, etc..), and bodies’ behaviors and actions. For the algorithm to become social software, in fact, ‘it must gains its power as a social or cultural artifact and process by means of a better and better accommodation to behaviors and bodies which happen on its outside.’ (Fuller 2008: 5).
Furthermore, as contemporary algorithms become increasingly exposed to larger and larger data sets (and in general to a growing entropy in the flow of data also known as Big Data), they are, according to Luciana Parisi, becoming something more then mere sets of instructions to be performed: ‘infinite amounts of information interfere with and re-program algorithmic procedures… and data produce alien rules.’ (Parisi 2013: X) It seems clear from this brief account, then, that algorithms are neither a homogeneous set of techniques nor they guarantee ‘the infallible execution of automated order and control’ (Parisi 2013: IX).
From the point of view of capitalism, however, algorithms are mainly a form of ‘fixed capital’, that is they are just means of production. They encode a certain quantity of social knowledge (abstracted from that elaborated by mathematicians, programmers, but also users’ activities), but they are not valuable per se. In the current economy, they are valuable only in as much as they allow for the conversion of such knowledge into exchange value (monetization) and its (exponentially increasing) accumulation (the titanic quasi-monopolies of the social Internet). In as much as they constitute fixed capital, algorithms such as Google’s Page Rank and Facebook’s Edgerank appear ‘as a presupposition against which the value-creating power of the individual labour capacity is an infinitesimal, vanishing magnitude’.( Marx 1973: 694) and that is why calls for individual retributions to users for their ‘free labor’ are misplaced. It is clear that for Marx what needs to be compensated is not the individual work of the user, but the much larger powers of social cooperation thus unleashed and that this compensation implies a profound transformation of the grip that the social relation that we call the capitalist economy has on society.
From the point of view of capital, then, algorithms are just fixed capital, that is means of production finalized to achieve an economic return, but that does not mean that, like all technologies and techniques, that is all that they are. Marx explicitly states that even as capital appropriates technology as the most effective form of the subsumption of labor, that does not mean that this is all that can be said about it. Its existence as machinery, he insists, is not ‘identical with its existence as capital… and therefore does not follow that subsumption under the social relation of capital is the most appropriate and ultimate social relation of production for the application of machinery.’ (Marx 1973: 699-700) It is then essential to remember that the instrumental value that algorithms have for capital does not exhaust the ‘value’ of technology in general and algorithms in partic