AI is not what you think! Everyday Life and the Digital Revolution
2nd May 2019
Silverstone Room, Pethick-Lawrence Tower, (PEL) - 7th floor, 16:00 – 17:30, followed by a reception
Speaker: Professor Anthony Elliott
Respondent: Professor Anthony Giddens & Professor Judy Wajcman
Chair: Professor Sonia Livingstone
In this lecture, Anthony Elliott argues that much of what passes for conventional wisdom about the AI Revolution is either ill-considered or plain wrong. The reason? AI is not so much about the future, but is rather a revolution already well underway – albeit one which is unfolding in complex and uneven ways across the globe.
From industrial robots to chatbots, and from driverless cars to military drones – AI is transforming all aspects of our lives, from the most intimate aspects of personal relationships to the changing nature of work, employment and unemployment.
Elliott focuses on the complex systems of AI - spanning intelligent machines, chatbots, advanced robotics, accelerating automation, big data - and their centrality to new forms of social interaction and interpersonal communication. In particular, the lecture examines AI in terms of the complex ways in which people communicate and talk with each other. Not only is machine-to-machine communication escalating today, but person-to-machine talk is also increasingly significant. Examining the rise of chatbots, softbots and virtual personal assistants, Elliott argues that AI is an increasingly integral part of how people live and work, now and in the future.
ANTHONY ELLIOTT is Research Professor of Sociology and Executive Director of the Hawke EU Jean Monnet Centre of Excellence at the University of South Australia, where he is Chancellery Dean of External Engagement. He is Super-Global Professor of Sociology (Visiting) at Keio University, Japan and Visiting Professor of Sociology at UCD, Ireland. He is a Fellow of the Academy of the Social Sciences in Australia. In 2018, he was appointed to the Expert Working Group of the Academy of the Council of Learned Academies in Australia to investigate: “Deployment of artificial intelligence and what it presents for Australia”. The project has been commissioned by the Chief Scientist of Australia at the request of the Prime Minister’s Commonwealth Science Council, and with support from the Australian Research Council (ARC), the Department of Prime Minister and Cabinet, and the Department of Industry, Innovation and Science. Professor Elliott is the author of many influential books in social theory and modern sociology, including most recently Reinvention, Identity Troubles and The Culture of AI, all published by Routledge.
Big Data and Social Ontology: When appearance Is calculation
Date: 1 February 2018
Venue: Silverstone Room, Tower 3, 7th Floor, LSE
Speakers: Professor Nick Couldry (LSE), Professor Jannis Kallinikos (LSE)
This talk will discuss the methodological – and more fundamentally, ontological – challenges for anything like social analysis when the ‘the social’, or at least particular important sites for sociality (‘social media’) are computed: that is, constituted by and through the outcomes of deep forms of data processing that are traversed by established instrumental practices of control and/or profit making. Interaction in social media is organized along highly stylized activity corridors (e.g. sharing, tagging, liking) that essentially serve the purpose of transforming the flows of online user participation into a computable and ultimately tradable data footprint. By these means, each user action is rendered a discrete data token, a measurable click-through. Discrete data are then aggregated and several scores of user choice affinities and relations are ceaselessly computed and fed back to users, thus establishing a dynamic context of interaction between user choice and sociality online. The talk will address these issues from the point of view of social ontology. ‘Ontology’ is itself an important topic, and increasingly so, in computing as well, recently, for data justice scholars. What can a social-theoretical approach add specifically here?
One approach to be discussed concerns the spaces of digital social platforms. All social spaces, not just public spaces, have until now been analysed on the basis that they are ‘spaces of appearances’ in Hannah Arendt’s term: we have assumed until now that ‘appearance’ in that space – presence-to-others, availability for interaction with and evaluation by others in that space – can be taken as a basic datum of social analysis. But what if ‘appearance’ is now the result of prior computer-based calculations, calculations driven by underlying economic motivations? What are the consequences of this for the reciprocal nature of the social that sociology and phenomenology have tended to assume?
Jennifer Gabrys, Goldsmiths, Sociology
Date: May 18th 2017
Venue: Silverstone room
In this presentation, Jennifer Gabrys considers how citizen sensing practices that monitor air pollution experiment with the tactics and arrangements of environmental data. These monitoring experiments, however, are not just a matter of enabling “citizens” to use technology to collect data that might allow them to augment scientific studies or to act on their environments. Rather, computational-sensing technologies are bound up with the generation of new milieus, relations, entities, occasions, and interpretive registers of sensing. Drawing on her recently published book, Program Earth, as well as material from the Citizen Sense research project, Gabrys will discuss how the becoming environmental of computation describes these sensing practices and processes. Sensor-based engagements with environments do not simply detect external phenomena to be reported; rather, they bring together and give rise to experiencing entities and thereby actualize new arrangements of environmental sensing and data. The production of air quality data through environmental monitoring generates distinct entities and occasions for generating and making sense of that data—as scientific facts, matters of concern, or even as inchoate patterns produced through unstable technologies or sporadic monitoring practices.
Machine Learning, Interested Epistemologies, and Economic Morality
Date: 26 January 2017
Venue: Silverstone room
Speaker: Dr Bernhard Rieder, University of Amsterdam
[Slides of the presentation]
This talk inquires into the social and political ramifications of algorithmic ordering techniques that are based on inductive modes of learning and argues that these forms of automated decision-making contribute to a spread of the epistemological stance Desrosières calls "accounting realism". Rather than focus on intentional discrimination, errors, or unintended biases, I will show how data analytics - and in particular machine learning - reveal and operate on the structured and unequal character of contemporary societies, extending the reach and applicability of theories of "human capital" (Becker 1964) and installing “economic morality” (Allen 2012) as a central normative principle. My goal is to situate recent technological developments in a series of longer trajectories that are linked together by a transition from universalist ways of thinking to purpose-oriented perspectivism. Machine learning, then, appears as both a deep and deeply interested way of knowing that challenges core tenets of liberal democracy.
From routine to repetition. Theoretical and methodological reflections on practice theory.
Date: 10 November 2016
Venue: Silverstone room
Speaker: Dr Hilmar Schaefer, Europa-Universität Viadrina, Frankfurt (Oder), Germany
The paper centres on the notion of repetition and takes it as the key concept of practice theory. The first part of the paper addresses a widespread bias towards stability and reproduction of the social in practice theory, points towards the need to take account of the dynamics of the social and develops an understanding of repetition, which draws on the poststructuralist positions of Jacques Derrida and Gilles Deleuze. The second part of the paper outlines three related dimensions of repetition. First, practices repeat themselves. A matrix of practices already exist before a subject comes into being and is continuously shaped in the process of taking up pre-existing practices. Second, practices are repeated by subjects. This dimension refers to the performances of competent bodies and points to the fact that practices need bodies in order to persist in time and space. Third, taking Derrida’s reflection on iterability into consideration, practices are also repeatable. While being in principle comprehensible to others, they are also at the same time susceptible to divergent interpretations, misunderstanding or change of meaning. This stresses the dynamics of repetition and the possibility of change. By thinking of practices in terms of repetitions that link different sites and instances, the methodology of practice theory is to follow the fragile relations which make up the (in)stability of the social, enabling it to grasp the specific contributions of bodies and material artefacts. This opens up sociological theory for analyses into the relationality and heterogeneity of the social as I intend to show in the conclusion of the paper.
Working to consume: consumers as the missing link in the division of labour
Date: 22 November 2016
Time: 3:00-4:30 pm
Venue: Silverstone room
Speaker: Prof Miriam Glucksmann, Professor, University of Essex.
Respondent: Prof Judy Wacjman, Professor, LSE
The labour associated with consumption is not new, but has been rapidly expanding in recent years as a consequence of both socio-economic change and technical innovation. Few goods or services are delivered ‘complete’ to consumers in the sense of being ready for use without further activity, yet the role of consumers in completing a system of provision is rarely acknowledged in theories of either work or consumption. This session argues that the work of consumers is a significant and constantly developing field of work, and proposes a conceptual framework for understanding ‘consumption work’ as part of the division of labour. Recognition of the interdependence between the work undertaken prior to and after the purchase of goods and services problematises any assumption that all post-purchase activity comprises consumption and calls for a conception of the division of labour that extends from the market and world of paid employment to encompass also the usually unpaid labour of the end user. I draw on current international comparative research (the work of food preparation, the household recycling of waste, and installation of domestic broadband) to explore the varieties of consumption work, their shaping by prevailing systems of provision, and their place within the division of labour.
Bend It Like Sardex: New Perspectives on Sustainable Development
Date: 28 April 2016
Time: 4-5:30 pm
Venue: NAB 2.08
Speaker: Dr Paolo Dini, Associate Professorial Research Fellow in the Department of Media and Communications
Sustainable development depends on a virtuous interaction between economic growth, strong democratic institutions, and technological innovation. These areas of agency depend to different extents on local and global interactions, and evolve at different time scales. In all cases the narrative of development has been dictated by the West for so long that “other" peoples who have followed alternative histories have little choice but to adapt and learn to speak it. The implication is that development carries with it a huge aggregation of power. One of the main challenges in development, therefore, lies in the difficulty to create a safe harbour from global credit shocks and different forms of speculation. Sardex, a B2B, complementary and electronic mutual credit system set up in Sardinia in 2010, has found a way to reinterpret locally the social construction that trumps all others: the creation of money. In other words, Sardex has been able to build such a sea-wall by creating a non-convertible mutual credit system in which the money-creation power is distributed to the circuit members. In this talk I will briefly describe how Sardex works and will analyse it from the sociological monetary theory perspective. In the space of electronic media, the trillions of dollars of daily global forex activity propagating through the global networks are to digital financial data like knowledge is to information, just a higher semantic layer that correlates with macroeconomic indicators. But as the scale of the system decreases the sociological dimension becomes increasingly important and the individual ever more visible. The challenges faced by Sardex, therefore, are huge. Not only does it need the bear the weight of the capitalist storm, but it must build strength internally by connecting the kernel of power in capitalism, one of whose ends they now hold, with institutional learning, good governance, and technological innovation.
Could Problems Take the Place of Knowledge in Digital Societies?
Date: 15 March 2016
Venue: The Silverstone Room, Floor 7, Tower 3, LSE
Speaker: Noortje Marres, Associate Professor and Research Director at the Centre for Interdisciplinary Methodologies, University of Warwick
In this talk, Noortje Marres will explore the claim that social inquiry has become problematic in digital societies. Digital ways of knowing society have been widely questioned in recent years, from critiques of online surveillance to the ever growing lists of tools and apps that have been withdrawn because of privacy problems and other allegations of ethical violations (Facebook search graph, girls near me, Samaritan radar, and so on). Strikingly, however, questions of epistemology tend to remain under-explored in these controversies about digital ways of knowing society: the question of whether digital devices allow us (or them) to know society in the way they claim to tends to be overshadowed by more urgent ethical, political and moral issues. Nevertheless, investigating this knowledge dimension is useful, this paper would like to propose, because it brings into focus a much wider potential transformation of digital societies, and digital social inquiry: in these societies, interactivity between social research and social life intensifies to the point that representational assumptions are thrown into crisis, not just intellectually, but publicly.
The Mediated Construction of Reality: Remembering Schutz and Elias
21 May 2015, Andreas Hepp (ZeMKI, University of Bremen) and Nick Couldry (Media and Communications, LSE)
In this presentation Hepp and Couldry outlined the project of their current book, The Mediated Construction of Reality (Polity 2016). The book will offer a critical reevaluation and rearticulation of the social constructivist ambitions of Berger and Luckmann’s 1966 book The Social Construction of Reality while revising an interest in the work of Norbert Elias, particularly his concept of figurations. Elias, they will argue, is a particularly important theorist on whom to draw in making social constructivism ready to face the deep embedding of the social world with digital technologies. Audio.