Platforms and Potency: Democracy and Collective Agency in the Age of Social Media

Abstract Attitudes to digital communication technologies since the 1990s have been characterized by waves of optimism and pessimism, as enthusiasts have highlighted their democratic and liberating potentials, while critics have pointed to the socially, politically and psychologically deleterious consequences of unchecked digital capitalism. This paper seeks to develop an analytical framework capable of appreciating and assessing the capacities of such technologies both to genuinely enhance democratic agency, and to become tools through which capitalist power is enhanced with widespread negative consequences. The paper in particular deploys my concept of ‘potent collectivity’ in order to name the type of democratic agency that such media technologies can be seen both to enable and enhance under certain circumstances, and to inhibit under others. It also considers the affective qualities of ‘potent collectivity’, and in particular the utility of a Deleuzo-Spinozan concept of ‘collective joy’ as designating the affective quality typical of ‘potent collectivity’. The paper uses the specific example of left-wing political activism in the UK during the period 2015-17 to illustrate the potential for platform technologies to enable new forms of democratic mobilization, while arguing for an analytical position that eschews any simple celebration of the liberating potential of new technologies; remaining sensitive to the negative features of ‘platform capitalism’.


Platforms and Potency
The story is often told that Bill Gates and Steve Jobs -for many years, the CEOs of Microsoft and Apple respectivelyboth restricted their teenage children's access to smartphones, computer games and tablets in the home (Ward). Apparently, they did so out of concern for the deleterious effects that they believed over-use of such devices to have on the mental health of children and young adults.
This anecdote is widely circulated, and resonates powerfully with many audiences, because it seems to crystallise a dilemma that faces any attempt to reflect on the specificity of the human condition in the twenty-first century: are digital media, social media and network culture essentially empowering or disempowering? Arguably, most attempts to consider their social and psychological effects can be understood in terms of this basic question. The extraordinary increase in communicative and information-processing capacity enabled by the 'cybernetic revolution' is undeniable (Gilbert,. But so is the vast increase in the power of states and corporations to surveil populations and individuals. At the same time, the apparently addictive, hypnotic or enervating quality of immersion either in virtual environments or in unceasing information-flows seems to have grave implications for the agency and autonomy of individual users.
In this article, I will offer a contribution to attempts to address this issue, not by offering any definitive answer to it, but by positing one useful criterion according to which the effects of such media and their uses may be judged. This criterion will draw on my 2014 book Common Ground: Democracy and Collectivity in an Age of Individualism, and in particular one concept which I propose there: 'potent collectivity'. As I will explain in more detail below, I use this term in the context of a general understanding of 'democracy' as designating any type of social, institutional, cultural or political practice that facilitates the emergence of collectivities on any scale that are capable of making and enacting shared decisions. In this article, in the context of a survey of historical and recent analyses of digital, social and platform media and their effects, I will show how the concept of 'potent collectivity' offers a potentially useful analytical tool with which to assess the political effects and affordances of such media.
This will not be an exhaustive survey of studies of the relationship between 'new' media, political organization and activism. Rather it will be a consideration of some political and philosophical criteria according to which the variable deployments, potential and dangers of such technologies might be assessed, from a specifically pro-democratic perspective. In particular, I will consider how the general tendency of platform technologies and 'platform capitalism' to facilitate massive aggregations of data, of communicative connections and of individual users (as distinct from the intensive fragmentation and cultural specialisation typical of 'post-Fordist' culture (Amin)) demonstrates a capacity both to 'potentiate' large, distributed, politically-efficacious collectivities; and yet also to subject populations to new forms of subordination and alienation.
The specific historic example with which I will be concerned here is the British Labour Party's 2017 general election campaign. This has been widely understood as a turning-point in the history of British political campaigning, characterised as it was by a recovery of mass participation in electoral campaigning and by a highly innovative use of mass social-media platforms and bespoke communication technologies to facilitate that participation (Tonge et al.). It is important to stress several points from the outset, however. One is that Labour did not win the election -although they increased their share of the national vote by a higher margin and in a shorter time than any other party since the introduction of full adult suffrage in the UK -while at the subsequent election, they were roundly defeated by Boris Johnson's pro-Brexit Conservative Party. Another is that this final defeat can be attributed largely to the relentless hostility of established media outlets to Corbyn and his supporters, as well as to the fact that their political opponents, between 2017 and 2019, had developed a range of successful counter-strategies in their own deployment of social media (Philo et al.). The very success of these counter-strategies lends considerable weight to the views of those media critics and theorists who see the development of massive online platforms as a largely or entirely regressive historical development, while the decisive role played by legacy media stands as a stark warning against any over-hasty attempt to characterise 'new' media as the only ones that matter Nonetheless, none of these factors undermine the observation that something remarkable happened in 2017, which is worthy of analysis and theorisation. It is therefore with that analysis and theorisation that this paper will be partially concerned.

Responses to the Digital Revolution
To contextualise my historical and theoretical arguments, I will first consider a range of political and critical responses that have been made in recent years to the rise and ubiquitous spread of social media platforms, and the technological protocols and infrastructures on which they are based. Since the 1990s, internet technologies have been widely seen as potential facilitators of new forms of collaborative culture (Benkler; Rheingold) or even as potentially new forms of digital 'commons' (democratically-accessible shared resources). There remain arguably strong grounds for optimism as to the democratic potentials of these technologies. At the same time, a recent wave of analysis has seen commentators such as Zuboff, Lovink, Davies, Couldry & Mejias and Seymour decrying the psychological, social and political effects of widespread social-media use. Here I will consider a range of approaches to the issues raised by these various perspectives, propose an analytical perspective capable of accommodating many of them, and consider some of the theoretical, analytical and political implications thereby raised.
Online culture first became widely visible in the 1990s. Since that time, successive waves of commentary have focused either on the progressive or emancipatory promise of digital communications technologies, or on the actual or potential harms caused by them. Although there have always been positive and negative reactions the internet and its affordances (Wellman and Hogan) , we can identify a loose pattern of waves of optimism and pessimism recurring since the mid 1990s, culminating in a wave of recent commentary that is best understood as expressing a radical ambivalence about the overall social, psychological cultural and political consequences of human culture having become thoroughly saturated, mediated and reconstituted by the ubiquitous deployment of digital communications networks.
In response to the optimistic, almost millenarian tone of early commentary on internet culture (Negroponte), from the middle of the 1990s a rising wave of critique pointed to the pro-capitalist bias, libertarian individualism and political naiveite of much early 'cyberculture' (Barbrook and Cameron) (Terranova). Notably, commentators pointed to the tension between this corporate individualism and the explicitly communitarian and community-building aspirations of early network culture (Turner).
The pessimism of such analysts seemed to be fully confirmed when the dot.com bubble burst in 2000, leaving many former participants in the technology sector (and their dependents) unemployed, bankrupt and thoroughly disillusioned (Cassidy). However, with the emergence of 'web 2.0' (O'Reilly) and the first social media platforms, much commentary expressed understandable enthusiasm about the potential for community-building, democratic mobilization and creative expression that this new technology might afford. Allied to the same libertarian Marxist tradition that had celebrated the 1994 Zapatista uprising most vociferously (Burbach et al.; Cleaver), Michael Hardt and Antonio Negri saw the new world of global communication as one in which the power of the self-organised 'multitude' would eventually outstrip that of capital and the state agencies allied to it (a complex of forces they named 'Empire') (Hardt and Negri, Empire; Hardt and Negri, Multitude).
This optimism seemed soon to be justified by the levelling effect of the first global social media platform, MySpace. on music culture: appearing to create a zone wherein global megastars and multinational corporations enjoyed no obvious advantage over amateur musicians and independent micro-labels (Gilbert, 'Capitalism, Creativity and the Crisis in the Music Industry'; Godwin), and by the generally disruptive and democratising effects of peerto-peer technology (Kostakis and Bauwens). But in fact, what was to follow was the establishment of streaming services (most notably Spotify and Apple Music) and YouTube as the primary channels for music distribution by the end of the following decade (Aguiar and Waldfogel), with the historically high disparity between the financial benefits accruing to corporations and content-producers remaining little changed since the exploitative days of the post-war music industry (Citi GPS; Taylor).
Similar democratic potential seemed to have been demonstrated in 2010-1, when uprisings against repressive governments across the Middle East were widely reported to have been co -ordinated largely through the use of social media platforms (Gerbaudo, Tweets and the Streets) (Treré). This phenomenon was echoed later in 2011 when large-scale social movements in Spain , Greece (Galis and Neumayer) and the US used the same platforms to co-ordinate massive and highly public demonstrations against the fiscal austerity that had been imposed on so many populations following the 2008 financial crisis. But the democratic optimism provoked by these events gave way to further disquiet as conservative government and authoritarian regimes took power from Spain to Egypt, and as the growth of gigantic platform corporations (Amazon, Facebook, Google, etc.), and revelations about their partial collusion with state agencies, seemed to usher in a new era of monopoly capitalism (Srnicek, Platform Capitalism) and ubiquitous panoptical surveillance (Zuboff). Once again, a disruptive expression of democratic potential seemed only to have given way to widespread consolidations of both state and capitalist power in new and alarmingly illiberal configurations. In the United States, the election of Donald Trump seems to have been at least partly assisted by the rise of the racist 'altright': a political formation existing almost entirely in the digital spaces YouTube, 4-chan and Reddit (Nagle).
The most recent wave of negative analysis of these phenomena has seen the rise of social media and the platform economy characterized as nothing less than an unmitigated disaster for human civilization and subjectivity. Commentators such as Shushana Zuboff and Couldry & Mejias have persuasively posited the growth of powerful platform corporations, engaged in ubiquitous and permanent data-collection, as threats to the very viability of 'autonomous' human subjectivity or any meaningful form of civil society . Political analysts have pointed to the role of social media platforms in weakening public trust in civic institutions and in promoting new forms of authoritarian populism (Davies; Gerbaudo, The Digital Party).
And yet very recent history affords apparent examples of new forms of democratic mobilization being directly enabled by the deployment of platform technologies. In Spain, for example, a number of significant new political formations have emerged from the context of the 2011 protests: both in the success of radical municipalist movements in Barcelona (Colau and Bookchin) and in the emergence and political influence of new radical-democratic party Podemos ('Between Movement and Party'). In a comparable vein, the 2017 UK General Election saw the Labour Party achieve a wholly unexpected increase in its vote share, led by Jeremy Corbyn, advocating for a radical social-democratic programme. This followed an influx of new members into the party and a mobilization of tens of thousands of them during the general election campaign that had no precedent since the 1950s. While actual numbers of party members (certainly) and activists (possibly; but this is almost impossible to ascertain) had been higher at previous moments in the intervening decades, there was no precedent during that time for the evident increase in those numbers. In the UK as elsewhere, declining party membership and mobilization was seen as a near-universal phenomenon of contemporary politics until very recently . The enabling of new, activist-led movements by the deployment of new media technologies is not a novel phenomenon. It played a key role in the global popularisation of the Zapatista movement in the 1990s(Burbach et al.), the grassroots activist culture that emerged, for example in the UK during that decade (Pickard) and the anti-globalization movement of the early 2000s Gilbert,Anticapitalism and Culture); and of course the connection between new media, the 'imagined communities' that they enable, and the and innovations in political thought and organisation goes back at least to the early modern period (Hill; Anderson). However, the mass entry of activists -often with backgrounds in these same movements -into mainstream party politics (Gerbaudo, The Mask and the Flag), represented a decisive break with several decades of perceived democratic decline.
Whatever one's partisan's affiliations, this democratic efflorescence and successful mobilization was undoubtedly an impressive achievement, that confounded almost all professional predictions. It is significant for our purposes because many qualified observers attributed this outcome at least in part to the creative deployment of platform technologies by key organisations and acting in support of Labour and Corbyn (Bell). These technologies were seen as enabling Labour activists to counter the extreme anti-Labour bias for the British media (Curran and Seaton) through the circulation of alternative propaganda and, crucially, through local mobilization and door-step campaigning, the organizational costs of which were drastically reduced by the implementation of mobile platform apps (for example, apps that enabled volunteers to identify specific localities in which to concentrate campaigning efforts on any given day). This account of a specific political mobilization presents a very different picture of the democratic and civic implications of platform technologies to that offered by extreme pessimists such as Zuboff. And yet it also presents no evidence with which to contradict their specific claims. Today, the extremely damaging and politically alarming consequences of the cybernetic revolution in general -and the rapid diffusion of social media in particular -seem to coexist and coincide with real and tangible manifestations of their long-cherished democratic promise. It is perhaps not surprising then, that some of the most thoughtful of the most recent commentary seem to occupy a space of radical and irresolvable ambivalence towards the costs and benefits of widespread social-media use (Lovink; Seymour), or makes deliberate efforts to avoid simplistically 'utopian' or 'dystopian' claims as to the political effects of new media technologies (Fenton; Treré).

Logics of Progress and Reaction
How is it that this latest stage of the cybernetic revolution seems to have produced such positive and negative effects simultaneously? And is it possible to synthesise the insights offered by all of these commentators and different historical experiences? It is possible to derive such an account from the history that we have outlined, if we take an approach that begins by acknowledging two key facts: First, it must acknowledge that innovations in computing technology (as in many other areas of scientific and technological development) are not just outcomes of corporate profit-seeking or pre-existing market demand, but emerge from complex social milieus in which state institutions, non-corporate researchers and low-level entrepreneurs all play significant roles (Levy; Mazzucato). This is not to dispute for a moment the enormous role played by venture capital, world-scale profit-seeking and, above all, the military-intelligence complex in the development of communications and information technologies since the early twentieth centuries (Levine); it is merely to note that these have never been the sole motive or creative force in driving their development. Secondly, it must have to take account of the genuinely empowering affordances of these technologies, particularly at the level of group-formation and co-ordination. For example, social media platform corporations may well survey and exploit their users, encouraging compulsive and addictive behaviours while commodifying their data and 'behavioural surplus' . But they are only able to do this to the extent that users participate in the platforms, and this participation cannot be understood without acknowledging the genuine opportunities that it offers for the manifestation of various novel forms of collective agency. In fact, as I will explain further below, I would argue that this is typical of innovative techno-economic institutions throughout the history of capitalism: such institutions necessarily constitute opportunities for the creativity and /or productivity of populations to be radically enhanced, even while those populations and their outputs are subject to intense regulation and to logics of commodification and exploitation.

Potent Collectivities in the Era of Platforms 1
The recent history of the UK Labour Party remains an instructive example when considering the specific case of novel platform technologies and their capacities to collectively empower (or disempower) users. When the campaign for the election of Labour's leader began in May 2015, very few observers or participants saw left-wing socialist Jeremy Corbyn as having any realistic chance of winning the contest. Social media was central to the development and momentum of the campaign to elect him, with specific Facebook groups and pages such as 'Red Labour' (Sellers) serving as crucial hubs for the circulation of propaganda, and for the mutual recognition of an otherwise disaggregated and dispersed population of potential Corbyn-supporters. A crucial point to understand here is the huge disparity that existed at the time between public perceptions and mainstream 'common-sense' as to the general distribution of opinion across the electorate, and the reality of that distribution of opinion as revealed by reliable but poorly-publicized quantitative survey data (Green and Prosser). On the one hand, the assumption that left-wing, anti-capitalist, social-democratic perspectives were, by the mid 2000s, entirely marginal to British public opinion was simply taken for granted as self-evident and uncontroversial by journalists, commentators and most of the voting public (Curran et al.) . On the other hand, all reliable survey data suggested that this was not true, and that such attitudes had remained consistently typical of 20-30% of the electorate (and considerably higher with regard to certain key areas of public policy) throughout the period of neoliberal hegemony initiated at the end of the 1970s (Gilbert, 'Corbynism and Its Futures'). One of the decisive effects of the social media was to make apparent to many of this disaggregated, demoralized left-wing public just how numerous, and so potentially influential, they actually were. It was quite possible, and quite normal in much of Britain throughout the 1990s, 2000s and 2010s, even for those themselves holding socialist or social-democratic political views, to regard such opinions as the preserve of a tiny, insignificant minority of the population. Access to social media demonstrated to many that in fact their opinions were shared by a far larger proportion of their fellow-citizens than they had previously assumed.
It is worth noting that this is by no means a historically unique or even unusual situation within the entire period of modern political history. The French sociologist Gabriel Tarde argued in the late nineteenth century that the emergence of the mediated 'public' -a dispersed collectivity capable of socio-political agency -was one of the key social phenomena of his time and of the coming age (Tarde). Tarde's formulation anticipates a number of later theories and commentaries. Benedict Anderson famously identifies the emergence of the modern idea of the nation with the mediation of 'imagined communities' by newspapers, as well as industrial-era transport and communications networks (Anderson). I have argued myself that democracy as such is always constituted by the existence and institutional enabling of 'potent collectivities' (Gilbert,Common Ground 47), of which Tarde's 'publics' might be regarded as the archetypical modern form. In this instance, for example, both the public that emerged in support of Corbyn's leadership of the Labour Party in 2015, and the activist movement mobilized by his leadership at the 2017 UK general election, were examples of 'potent collectivities' enabled directly by the deployment of platform technologies. Significantly, the very possibility of Corbyn's election as Labour leader by a mass influx of members and supporters into the party had itself been institutionally enabled by a change in the party's leadership election rules in 2014 permitting direct election by the membership (rather than election be an electoral college representing local parties, trade union affiliates and Members of Parliament) (Nunns): a practice encouraged and enabled by the technological possibility of direct online voting (which significantly reduced the cost and logistical challenges posed by such a mass direct election).
These formulations can shed some light on issues raised by the emergence of the pro-Corbyn movement and the mobilisation of activists up and including the 2017 election. Although Tarde's 'public' and Anderson's 'imagined community' might be seen as exemplary forms of modern, mediated collective entity, both arguably also tend to be char-acterised by high levels of homogeneity and an organisational dependence on highly centralised media: most notably, mass-circulation newspapers. The pro-Corbyn movement was widely criticised during the period 2015-17 as a mere personality cult (Freedland; Cohen), the members of which would never actively participate in any meaningful political activity. However, Labour's unexpected success during the 2017 election was at least partly attributable to the fact that mobile and platform technologies enabled a very high level of distributed, decentralised self-organisation, with campaigners being actively encouraged to engage with voters directly and creatively, rather than merely reproducing centrally-determined campaign messages (in stark contrast to the Labour Party's traditional reluctance to encourage any degree of autonomy on the part of campaigning supporters). In this sense, the 'potency' of the collectivities enabled by these technologies, in this instance at least, was as much as anything a consequence of their departure from established modes of public and collective engagement more typical of nineteenth and twentieth century political culture.
The idea that social media platforms, like earlier iterations of digital communications technology, facilitate the emergence of 'potent collectivities', provides a useful formulation with which to make sense of the history under investigation here and the analytical questions raised by it. This concept is partly derived from the tradition of 'autonomist' Marxist thinking (such as Hardt and Negri) on the relationship between capitalism and the forms of collective agency that it enables, depends-on and suppresses at different moments. In brief, this account maintains that all forms of capitalism depend in part upon the organization of populations into collectivities that are technologically and organizationally enabled to increase their productive and creative capacities: collectivities that must then be subject to appropriate forms of regulation and administration in order to extract surplus value from them. This is true of the workers in a factory, the inhabitants of a commercial city, or the participants in a social-media platform. Crucially, in each of these instances, the fact that the population in question cannot be relied upon spontaneously to exhibit the necessary behaviours to generate surplus-value is fundamental to understanding the political situation of these 'potent collectivities'. The labour and socialist movements that emerged over the course of the nineteenth century were entirely predicated on the recognition that the potency inhering in industrial collectivities could be directed to other ends than simply maximizing profits for capital, if workers were sufficiently organized, educated and motivated. There is no question that a strong element of economic compulsion drove workers into the first industrial factories (Hobsbawm 200-07), as social necessity drives many users onto social-media platforms. But there was never any way for capital to generate profits without potentiating the collectivities that it helped to bring into being inside them. Today, the users of social media are clearly not merely tricked or pressured into participation and the surrender of profitable data; they participate to the extent that they find themselves actually enabled to form productive and useful connections with others through their use. Indeed, from the point of platform corporations, the most valuable behavioural data is only generated by users to the extent that their online behaviour generates detailed and accurate information about their real preferences and responses (Zuboff). Without users persistently engaging in meaningful and potentially significant interactions, this information would not be available to harvest. And yet: harvested it must be.
I would suggest that this formulation provides a coherent and comprehensive explanation of the logic of innovation, potentiation, capture and commodification that has characterized the relationship between users of social media (and their antecedent internet technologies) and the corporations (and even state institutions) that have used them significantly to enhance their power. From this perspective, the waves of commentary that we have outlined have all responded perceptively to real shifts and developments in the long political history of the cybernetic revolution. At each successive 'stage' (and of course any designation of coherent and discrete 'stages' in such a process is purely arbitrary and provisional), genuinely 'potent collectivities' have been enabled to emerge and have subsequently been subject to mechanisms intended to inhibit the collective autonomy of populations of users while directing their behaviour towards acts of individual consumption or politically-disabling forms of neurotic, narcissistic and / or addictive behaviour (Hearn; Seymour).
From this perspective, a key issue regarding social media and other digital platforms is the extent to which they do or do not enable the emergence of 'potent collectivities'. This term designates any social formation or group on any scale that is capable of producing positive effects in the world; of collectively taking and acting upon political, social or aesthetic decisions; or of constituting sites for innovation and invention. As I argue in my book Common Ground, democracy as such can be understood as the institutionalization of sites for the emergence of 'potent collectivities', at which the inherent complexity and multiplicity that characterise all groups (and indeed all phenomena on any scale) can find political, social or aesthetic expression (Gilbert 2014).

What Makes a Collectivity Potent?
What criteria might be applied for the identification of such 'potent collectivities'? On a certain scale, almost any form of successful activity enabled by a social media platform might be seen as exemplifying the constitution of a 'potent collectivity': a group of friends or neighbours organizing a small house-party, for example. On one level this itself is a reasonable and politically significant claim. As I have argued (and many others have before me), regimes of power (including, but not limited to the regime of neoliberal hegemony) often operate through the deliberate, active and institutionalized inhibition of productive social relations between subjects (Gilbert, Common Ground 47) from the criminalization of labour organization to the promotion of artificial relations of market competition within public services (Leys) , to the segmentation and physical isolation of prisoners described so famously by Foucault (Foucault). I follow Protevi, Deleuze and Spinoza in characterizing as 'joyous' social relations and experiences wherein bodies experience an enhancement of their capacities and opportunities for production relationships with others, and the production of such relations even on a small scale can surely be seen as marking some kind of democratic advance even on micro-social scales. To offer an isolated autobiographical example, at my own children's primary school in East London, the habit of parents of children within a particular class using WhatsApp groups to share useful information, organize emergency pick-ups and childcare, and even to discuss serious problems in the social relationships of the children has facilitated forms of useful communication and collaborative care that were far more difficult to implement before the popularization of the platform. Of course, this could itself be interpreted as merely substituting for more traditional or institutionalized forms of community and collaboration that have been displaced by the increasingly individualized patterns of work and sociality typical of advanced neoliberal culture, and as ultimately benefitting the employers of hard-pressed parents more than anyone else (because it effectively enables them to exert slightly more pressure on parents to informally extend their working hours, knowing that they will be able to use the groups to arrange for emergency care for their children at short notice). Nonetheless the potential, on however small a scale, for forms of collective agency to be enabled by platform media, is clearly indicated by this example as well as by formal research into parents' use of social media to constitute support networks (Cole et al.).
Such microsocial illustrations are useful, but I offer them here only as a step on the way to clarifying the concept of 'potent collectivity'. At this stage in the analysis, it seems difficult to say whether the WhatsApp groups in question are actually facilitating 'potent collectivity' or merely enabling social reproduction under conditions of general exploitation and alienation (Luxton and Bezanson). To move beyond this ambiguity, and to establish some reasonable criteria for the identification of 'potent collectivities', we might simply say that 'potent collectivities' are characterized by at least some capacity to overcome obstacles, opposition or antagonism from hostile forces. This would be consistent with an understanding of political identity as in part always constituted by antagonistic social relations, and political agency as in part always constituted by the capacity to take and enforce decisions, both perspectives being common to a number of well-established theoretical currents (Laclau).
From this perspective, the groups of technologically-enabled activists studied by analysts such as Hands and Gerbaudo clearly manifest more obvious features of 'potent collectivity', having found myriad creative ways to use internet technologies to facilitate self-organisation and collective decision-making (Maeckelbergh). And here again, the example of the UK Labour Party's 2017 General Election campaign remains instructive. Many commentators and analysts over the years have pointed to the tensions between forms of networked, 'horizontal', deliberative and radically-democratic self-organisation on the one hand, and the exigencies of strategic goal-oriented political organisation in the face of a determined enemy, on the other (Mouchard). What is particularly instructive about this example is that it clearly constituted both an antagonistic mobilisation against a clearly-defined and well-resourced set of enemies and an unusually high level of self-organised and autonomous media-production for a modern electoral campaign. These enemies included the Conservative Party, its allies in the press and wider social elite; but also the right-wing of the labour party and the trade union bureaucracy, who were widely understood to regard a positive result for the party at that election as both impossible and undesirable (because it would cement the leadership of Corbyn, which they saw as a temporary but insupportable obstacle to their continued control of the party) (SKWAWKBOX). It was this latter element of the situation that more or less necessitated high levels of self-organisation by the hundreds of thousands of new labour activists inspired by Corbyn's leadership, many of who were largely excluded from participation in ordinary campaigning activities by the reluctance of the party bureaucracy to organise them or induct them into the organisational life of the party. The response to this situation was made primarily by Momentum, an organisation established to support Corbyn's leadership of the party, whose main organisational tool was its access to the database of hundreds of thousands of party members and supporters to have registered support for that leadership during the leadership election campaign of 2015 ('What next for Momentum -and Might the Tories Beat Them at Their Own Game?'). Organised primarily through mass emails, Momentum members also used an innovative suite of custom-designed networking applications, allowing them both to identify and travel to priority campaigning areas to canvass local households and to maintain contact with voters they had met through a unique messaging app. Both the official Labour campaign and a network of independent supporters and alternative news outlets (e.g. the Canary: https://www.thecanary.co) used Facebook and YouTube to generate and disseminate propaganda reaching hundreds of thousands of voters at negligible cost (Savage and Hacillo).

The Power of Aggregation
The effectiveness of this mobilisation was in some senses unsurprising, given that historically Labour's greatest political weakness had been the near monopoly of support among commercial media outlets enjoyed by its Conservative opponents since the 1970s (Curran and Seaton). At the same time, it seemed to illustrate the potential of social media to make certain key potentialities available to social actors other than giant platform monopolies. The fundamental capacity of platform technologies, on which 'platform capitalism' depends, is arguably their capacity for sheer massive aggregation: of data, of people, of relationships and of objects. In this sense, the immanent social logic of 'platform capitalism' is quite different from that of 'Post-Fordism' (Goffey). Post-Fordism was often characterised as producing and relying upon an ever-more specialised differentiation of consumer markets, of electorates and of populations in general. It was this logic that was seen as producing the social and cultural logic of fragmentation that characterised 'postmodern' culture and society. While it is clear that on certain scales and at certain scales, platform capitalism merely intensifies this logic of fragmentation and dislocation (Davies), it is also clear that it also has (and partly depends upon) a capacity to aggregate and co-ordinate large numbers of people, ideas and political demands Williams). The astonishing recovery of Labour as a mass party, at a time when many commentators still assumed that the era of mass parties was long passed (Bell), serves as an invaluable illustration of this principle. It is worth reflecting, then, that this capacity for aggregation was arguably one of the features of the cybernetic world discerned by Hardt and Negri in their optimistic diagnoses of the prospects for collectivist politics at the turn of the twenty-first century. Indeed, it is worth noting that one of Hardt & Negri's main intellectual influences -Félix Guattari -had himself discerned great promise in the political potential of networked media, as far back as the early 1980s (Prince and Videcoq).

The Power of 'Collective Joy'
Another factor for consideration, that is highly relevant for our purposes, is the affective specificity of Labour's 2017 campaign as experienced by activists and by large sections of the wider public. A sense of elation, empowerment and liberation, all connected with a sense of genuine collective agency and an intensively enhanced capacity for individual activists to communicate both with each other and with voters, were routinely reported by activists involved at every level of campaigning (Savage and Hacillo). This manifestation of the experience of 'potent collectivity' exemplifies both the undeniable potential of platforms media to enable such experiences, and their centrality to processes of political mobilisation in general.
It is a very widespread assumption of progressive and radical politics that what provokes subordinated, disempowered or exploited populations into mobilisation for democratic objectives is the accumulation of grievances past a certain threshold of tolerability. It is clearly outside the scope of this article to contest this assumption authoritatively. What can be stated is that from the perspective of the theoretical traditions on which the concept of 'potent collectivity' draws, it is mistaken. From this perspective, it is never merely a consciousness of oppression that motivates such mobilisation, but always of a certain collective potency and potentiality: 'joyous affect', to use the terms that John Protevi derives from his readings of Spinoza and Deleuze (Protevi 51). From this perspective, positive affects are always associated with an awareness of enhanced potency, negative affects with a sense of reduced potency. Enslaved populations have endured abject slavery for generations, while depressed and alienated individuals rarely (arguably never) convert those affects into radical political activity (Fisher). In the case of the sequence of mobilisations that we have been discussing here, it was the 'joyous' realisation of their immanent potential actually to affect the outcome of a Labour leadership election that first mobilised the 'potent collectivity' of several hundred thousand activists that would both win that election for Corbyn and achieve a historic result at the 2017 General Election.
The importance of this type of 'passionate' political engagement to mediated political processes has been usefully highlighted by theorists such as Natalie Fenton (Fenton 81-103). But it is also worth specifying that the Spinozan concept of 'joy' does not necessarily involve any experience of elation, ecstasy or even hopefulness. As Frédéric Lordon writes: "The concept of 'joy' must be understood here with a certain analytical coldness, emptied of the ideas of rapture, plenitude or jubilation that are commonly associated with it. One can experience joy at all levels of intensity, including very low ones, associated with the most ordinary; it can even go unnoticed, lost within a larger complex of affects that makes it hard to isolate." (Lordon and Ash 29).
At the same time, the specific details of the form that mobilisation took are of particular interest here, because they offer some crucial insights into that the specific qualities of 'potent collectivities' may be, and how they may be differentiated from what we might call other modes of relationality (Gilbert,Common Ground 129) that social media platforms can encourage. In the specific context of the 2017 Labour mobilisation, we seem to find a concrete instantiation of some of the principles we hinted at when considering the history of digital platforms as empowering technologies. What was facilitated by this set of organisational techniques and the platform applications that enabled them, was a range of communicative relations characterised by a high-level of internal complexity and a low level of predetermination. Activists were able to produce, generate, circulate and modify a range of messages and interventions on multiple scales: from the mass-participation platform to the individual 'persuasive conversation' on the campaigning doorstep (Elgot). Productive relationships between campaigners, and between campaigners and voters, could be generated and managed effectively. The precise content and nature of these relationships was largely determinable by the participants in real time, according to an improvisational logic of social creativity. This capacity to facilitate invention and collective experimentation on the part of users is arguably a crucial feature of any political or media apparatus -or indeed any policy regime -that is truly pro-democratic in nature (Gilbert, 'The Joy of Co-production').

Sad By Design
We can usefully contrast this set of characteristics with those that social critics identify as typical of massive social media platforms. A repeated criticism of such platforms is their reliance on the 'like' button to reduce all interaction between users and messages to a one-dimensional, quantifiable indication of generic approval or its absence. A growing body of literature shows how this device has been deployed specifically in order to encourage compulsive behaviour in users, with the 'variable reward' of receiving an unpredictable number of 'likes' for any post or comment compared directly to the dopamine-manipulating psychotechnics of electronic gambling machines (Seymour). Here, users seem to experience their relationship to the general field of relations and communication enabled by the platform in terms which are, indissolubly, both pathological and individualised. Rather than the 'joyous affect' of open-ended lateral relationship-building, what seems to ensue is a sense of 'machinic enslavement' (Lazzarato and Jordan 12-17) whereby the user is compelled to engage with the system monotonously, repetitively, but entirely unproductively. At the same time, while the 'like' button might seem at first to be an innocuous way to allowing users to connect with each other, it is very well-documented that many users experience participation in the platforms as a permanent competition (both with other users and with themselves) for 'likes'. This imposition of competitive social relations within a social field that is not necessarily characterised by their preponderance is, according to many commentators, absolutely typical of neoliberal culture (Davies and Chakrabortty).
We could make an analogy here with the way that various thinkers have conceptualised types of relationships between component members of groups. In his classic study of internet-enabled activism, Joss Hands deploys Jean-Paul Sartre's distinction between the 'serial mode of co-existence'  of populations whose members have no direct or meaningful relationship to each other, and the 'fused group' capable of actual collective agency. This correlates interestingly with my claim that neoliberalism (and capitalist culture in general) tends to promote 'leviathan logic' (Gilbert,, according to which groups are only held together by the singular 'vertical' bond between each individual and some central locus of identification or leadership. The hyper-individualised relationship of the alienated social-media addict to the broader network can be understood here as a contemporary manifestation of serial leviathan logic: the wider socius enabled, actualised and manifested by the platform can only be connected-to via the straightened, affectively impoverished mechanism of the like-button, producing a relationship of irremediable dissatisfaction and a permanent reduction of agency. The pathological relationships increasingly reported by social-media users in terms of 'addiction' to platforms and their variable rewards can thus be understood as a very specific, but also entirely typical, manifestation of certain key features of our current phase in the history of capitalism. The giant platform corporations -on some measures the wealthiest and most successful institutions in the history of capitalism -are only able to succeed by enabling the genuine potentiation of collectivities on multiple scales, and the facilitation of 'joyous' collective agency and 'joyous affect' in the process. At the same time, they are only able to use these potentiated collectivities as sources of massive surplus-value to the extent that they subject them to mechanisms of surveillance and behavioural manipulation that have the effect of privatising and individualising the experience of many users, producing debilitating rather than empowering effects and thereby producing those clusters of negative affects that are so often understood in terms of depression and addiction. It is this logic that leads to the production of techno-social assemblages that seem to be, in the words of media critic Geert Lovink, 'sad by design'. Lovink points to the prevalence of 'sad' affects within social media culture, as users experience and express high levels of depression and alienation refs (Lovink). 'Sad' affects are, in the Spinozan tradition, the precise opposite of 'joyous affect' (Lordon and Ash 23-28): they are what is experienced when the capacity of bodies to relate to other bodies -and consequently to act in the world at all -is diminished. As Lovink writes: The hollow ache of sadness hurts. Yet it's also important to locate it empirically, to investigate its specific conditions. Far from being a natural response, such sadness is integrated into the design of interfaces and the architectures of apps. Today sadness has become technical, a shift that applies equally to users and producers. (Lovink 51) From the perspective being developed here, such sadness can be understood as a predictable product of capitalist deployments of platform technologies. In their impressive recent account of the political implications of digital capitalism, Nick Couldry and Ulises A. Mejias argue that the new mechanisms of platform monopoly are so invasive and generate such a complete imbalance of power between individual users and the corporations who harvest their data that the only appropriate historical analogy is with the early phases of colonialism, when entire populations were rapidly subjugated and deprived of even the most rudimentary elements of autonomy. For Couldry and Mejias, the best historical analogue of the types of collectivity brought into being by platform technologies would not be the potentially-militant industrial factory, but the slave plantation . If we further consider, for example, the recent proliferation of obfuscatory conspiracy theories via social media (Rosenberg and Steinhauer), it is clear that platform media have an extraordinary capacity to render populations and individuals collectively unable to effect meaningful engagements with social reality, subjecting them to novel forms of 'machinic enslavement' (Deleuze and Guattari, A Thousand Plateaus 492; Lazzarato and Jordan) of which the sadness described by Lovink would be only one localised symptom. 'Impotent collectivity' might be the ideal term for the type of distributed and generalised passivity engendered by such mechanisms.

Platform Capitalism, Platform post-Capitalism 2
My overall analysis here has a number of conceptual and political implications. On the one hand, it would seem to validate many of the perspectives to be derived from a particular tradition of critical thought including Deleuze & Guattari, Hardt & Negri, Lazzarato, Jason Read, etc. (Deleuze and Guattari, Anti-Oedipus; Deleuze and Guattari, A Thousand Plateaus; Hardt and Negri, Empire; Read, The Micro-Politics of Capital; Read, The Politics of Transindividuality). For example, Seymour shows that compulsive and addictive relationships to social media seem to exhibit classic symptoms of the 'compulsion to repeat' that Freud identified as fundamental to human psychic life in his analysis of the 'death drive' as one of the fundamental 'drives' constituting the psyche . However (and nothing in Seymour's analysis would contradict this observation), Deleuze and Guattari posit such 'fundamental' elements of human subjectivity identified by psychoanalysis as socially and historically specific features of the forms of subjectivity produced by capitalism: particularly focussing on the artificial production of 'lack' in 'oedipalised' subjects (Deleuze and Guattari, Anti-Oedipus). The deliberately-addictive deployment of the 'like' button would seem to be a perfect instantiaton of such a phenomenon here (Eyal).
Further, the specific forms of surveillance and behavioural modification promoted by platform capitalism seem to manifest perfectly the logic of 'control societies' first identified by Deleuze in 1990 (Deleuze). In his famous essay on this subject, Deleuze sees emerging a new world of networked power in which state and corporate agencies will offer as much freedom of possible to individual actors within a general field of social activity, rather than imposing centrally-determined norms on populations, but will seek to constantly modify their own institutional mechanisms in order to anticipate and profit from the likely behaviour of individuals and groups on multiple scales. There could hardly have been a more accurate anticipation of the behaviour or platform monopolies, whose business model is entirely dependent on the collection of 'behavioural surplus': data about the behaviours and preferences of platform-users that can be sold-on for marketing purposes. The 'like' button, it must be remembered, does not only serve to promote compulsive behaviour among those users (thereby keeping them engaged with the platform, thereby generating behavioural surplus): it also offers a very reliable mechanism for the surveillance, aggregation and quantification of user preferences.
At the same time, Hardt and Negri's millennial optimism has clearly not been comprehensively disappointed, and their understanding of what specific characteristics features might characterise 'potent collectivities' (although that is not the term they use) remain highly relevant. For Hardt and Negri, it is the internal complexity, non-linearity and distributed nature of 'the multitude' that are indissociable from its creative and communicative capacity. The experience of instances such as Labour's 2017 election campaign seems to bear out the implication that a high level of self-organisation and 'horizontal' relationality characterise situations in which 'potent collectivities' can be effectively enabled by the creative deployment of platform technologies. The internet can be the mechanism through which 'commons'shared resources characterised available to entire populations of users (M De Angelis) -and 'potent collectivities' are built.
But clearly, it is not necessarily that. The final conceptual observation to make here is that any comprehensive account of the politics of social media platforms and the culture that they produce must take account of their negative consequences; and that such an accounting seems to require some attention to the specifically individualising and privatising mechanisms that have been deployed by those platforms in recent years. Here, what seems to be in effect is a specific interaction between the techno-social potential of a radical new technology (social media), the socio-economic imperative of capitalism both to potentiate collectivities and to exploit them for profit, and the tendency of hegemonic neoliberalism to promote privatised and competitive relations in every social field. Without understanding the importance of each of these elements and of their complex interactions, we cannot understand the complex and ambivalent politics of contemporary social-media culture.
Finally, what might be some of the political implications of this analysis? The first is to note the widest possible application of some of the conceptual claims that I have made and illustrated here. The fact that capitalism always operates partially through the potentiation of collectivities, and must always act upon those 'potent collectivities' to prevent the manifestation of their latent democratic potentials, is an observation with significant political and analytical implications. And the cases that I have referred to here seem to be bear out this claim quite clearly.
The second is to note that there appear to be clear limits to how far platforms designed entirely by profit-seeking corporations can be expected to facilitate the emergence of 'potent collectivities'. Labour's 2017 campaign could not have taken the form it did without access to those platforms, but it also could not have done so without dedicated apps designed and deployed by its own technicians. At the same time, it would not have been able to make this deployment successfully without access to the institutional resources and real-world culture of a major, organic institution: the Labour Party itself. Many of the activist mobilisations studied by Hands and Gerbaudo, and even the new wave of 'digital parties' (such as Podemos in Spain and the Italian 'Five-Star' movement) (Gerbaudo, The Digital Party), have suffered from the lack of such organic institutional bases, and so seem to have found it very difficult to creative stable, lasting and positive sets of socio-political relations that do not reproduce the problematic social logics of corporate platforms: individualising members, promoting shallow forms of democracy, over-intensifying the power of isolated celebrity figures. What seems to have been most productive in the case of Labour 2017 is the creative interaction between deployments of platform technologies and far more traditional and 'analogue' forms of political mobilisation, such as face-to-face local campaigning.
It is highly likely, given the history that we have examined here, that a key political issue for the foreseeable future will be the relationship between state institutions and social-media platforms. Seymour concludes his study of social-media culture by reflecting on the lost potential of publicly-owned networks such as France's Minitel (a government-funded national videotext network that existed between 1980 and 2012) . Commentators such as Srnicek have called for the nationalistaion of platforms (Srnicek, 'The Only Way to Rein in Big Tech Is to Treat Them as a Public Service | Nick Srnicek'). As we have seen, governments have been heavily involved in funding and administering both fundamental research and vital infrastructure throughout the history of the cybernetic revolution. It seems highly unlikely today that the most negative consequences of that revolution -such as the ubiquity of technologies deliberately designed to be addictive to users -can be avoided without removing or severely curtailing the profit motive that has driven the platform monopolies into the transparently destructive forms of behaviour that Couldry and Mejias justifiably call 'data colonialism' (Couldry and Mejias).
As such, perhaps the most useful and vital implication of these reflections is to offer some guidance as to what progressive political demands might constitute in relation to the still-developing platform economy and the ongoing cybernetic revolution. Massive online platforms, it is clear, can function both as unprecedentedly effective sites for the actualisation of 'potent collectivities', and as sites for the extraordinary disempowerment, alienation and exploitation of individuals. Perhaps a key democratic demand for the years to come should be to ensure that the former potentiality is expressed at the expense of the latter. As we have seen, this is likely to involve the demand for technologies and applications capable of facilitating horizontal relations, open-ended relationality and sustainable, creative institution-building. This is not a demand that is ever likely to be met by profit-oriented corporations. As such, it is only likely to be achieved if platforms can be transitioned from their current legal-economic status into new ones that effectively institutionalise their status as democratic commons. There is no technical obstacle whatsoever to Facebook, Google, Youtube, and even Amazon being taken out of the hands of shareholders and transformed into massive, self-managing 'platforms co-operatives' (Scholz and Schneider). No lesser objective is likely to enable to realise their potential for democratic potency, or to minimise their capacity for destructive and alienating harm.