RGS-IBG Annual International Conference 2019

RGS-IBG Logo
Add to my calendar:    Outlook   Google   Hotmail/Outlook.com   iPhone/iPad   iCal (.ics)

Please note that some mobile devices may require third party apps to add appointments to your calendar


235 Geographies of/with Artificial intelligence (1): Spacings
Affiliation Digital Geographies Research Group
Convenor(s) Sam Kinsley (University of Exeter, UK)
Chair(s) Sam Kinsley (University of Exeter, UK)
Timetable Thursday 29 August 2019, Session 3 (14:40 - 16:20)
Room Sherfield/SALC Building, Room 7
Session abstract We are variously being invited to believe that (mostly Global North, Western) societies are in the cusp, or early stages, of another industrial revolution led by “Artificial Intelligence” - as many popular books (e.g. Brynjolfsson and McAfee 2014) and reports from governments and management consultancies alike will attest (e.g. PWC 2018, UK POST 2016). The goal of this session is to bring together a discussion explicitly focusing on the ways in which geographers already study (with) ‘Artificial Intelligence’ and to, perhaps, outline ways in which we might contribute to wider debates concerning ‘AI’. There is widespread, inter-disciplinary analysis of ‘AI’ from a variety of perspective, from embedded systematic bias (Eubanks 2017, Noble 2018) to the kinds of under-examined rationales and work through which such systems emerge (Adam 1998 Collins 1993) and further to the sorts of ethical-moral frameworks that we should apply to such technologies (Gunkel 2012, Vallor 2016). In similar, if somewhat divergent ways, geographers have variously been interested in the kinds of (apparently) autonomous algorithms or sociotechnical systems are integrated into decision-making processes (Amoore 2013); encounters with apparently autonomous ‘bots’ (Cockayne et al. 2017); the integration of AI techniques into spatial analysis (Openshaw & Openshaw 1997); and the processing of ‘big’ data in order to discern things about, or control, people (Leszczynski 2015). These conversations appear, in conference proceedings and academic outputs, to rarely converge, nevertheless there are many ways in which geographical research does and can continue to contribute to these contemporary concerns. This session aims to make explicit the ways in which geographers are (already) contributing to research on and with ‘AI’, to identify research questions that are (perhaps) uniquely geographical in relation to AI, and to thereby advance wider inter-disciplinary debates concerning ‘AI’.
Linked Sessions Geographies of/with Artificial intelligence (2): Working
Contact the conference organisers to request a change to session or paper details: ac2019@rgs.org
Automating production of the built urban environment: tracing the role of BIM modelling, VR and automated prefabrication in the UK housing sector
Rachel Macrorie (University of Sheffield, UK)
Computational intelligence is heralded as the means to revolutionise the intransigent UK housing construction industry by predicting better results, enabling coordination of complex projects, and scheduling repetitive decision-making tasks. Increasingly, Building Information Modelling (BIM) systems, which allow almost anytime, anywhere access to project data throughout the building construction lifecycle – are being adopted by the sector. Their adoption is supported by the Digital Built Britain programme, which promises to ‘better understand the needs of users and enable ‘right first-time delivery’, ultimately providing buildings and infrastructures as quickly and efficiently as possible’ and with enhanced levels of control (BEIS and Innovate UK, 2017). Coupled with augmented/virtual reality, next generation 6D BIM promises a ‘flawless futuristic construction’ digitising the whole construction process through real-time visualisation (NIBT, 2018). Taking the case of the sustainable housing developer CITU and their pioneering Climate Innovation District in Leeds (West Yorkshire, UK) (CITU, 2018), this article traces production of a low-energy smart home from conception and design in a virtual reality (VR) showroom, through part-automated prefabrication and onsite assembly, to digitally-monitored occupancy and maintenance. This production lifecycle is managed using BIM. Along this route I analyse: first, how is this experiment an attempt to produce a systemic change in the socio-technical infrastructure of housing and its construction? Second, how are socio-material ways of knowing, practices and processes (re)configured by automation and AI technologies, and with what implications? Third, what are the opportunities and limitations of automation and AI in this sector, and with what consequences for energy transitions?
AI/Machine Learning algorithms in public participation
Yu-Shan Tseng (Durham University, UK)
AI/Machine Learning algorithms are increasingly making consequential decisions, both in our everyday lives (e.g. Facebook/Netflix/Spotify) and with regard to how we are governed (e.g. in the use of algorithms in border control systems) (Amoore 2013). However, we remain largely unconscious of which decisions AI/Machine Learning algorithms are making for us (Beer 2009; Thrift 2004), and of the consequences and political implications of AI. Will AI improve democracy, or lead to further political polarisation? This research focuses on the application of AI/Machine learning algorithms in a public participatory process facilitated via open-source software (Pol.is) embedded on the vTaiwan platform. Adapting machine-learning algorithms (Principle Component Analysis and K-means clustering algorithms) from Netflix and Facebook, Pol.is aims at creating a participatory landscape where Machine-Learning algorithms (rather than politicians or moderators) create different ‘opinion groups’ and select ‘common opinion’ between them. I view Machine Learning algorithms through the lens of infrastructures. Working as infrastructures, they have configured participatory processes into standardized yet abstract, multidimensional and Machine-readable relationalities (i.e. particular relationships between participants and comments, and correlations between different participatory patterns). They translate abstract relationalities into comprehensible diagrams which visualise opinion groups and common opinions. Potentially, it could be argued that algorithms have imposed a new ontology in public participation through the provision and calculation of particular relationalities. Viewing them as digital infrastructures also allows us to question in terms of their visibility for participants (to what degree they should be visible to participants) and political consequences (are these algorithmic-calculated results making changes in policy-making)?
The algorithmic production of space in the age of machine learning: the case of self-driving vehicles
Fabio Iapaolo (Polytechnic University of Turin, Italy)
From speech recognition to self-driving cars, contemporary automated systems rely on learning algorithms, that is, software programs that can learn from example tasks so as to become gradually independent at producing a certain output. The widespread use of machine learning algorithms to automate mundane tasks has introduced an ambiguous conceptual overlap between automation and autonomy. This is because learning algorithms do not simply execute rule-based instructions. Rather, they can adaptively operate within ever-changing environments, taking decisions with no human input and beyond predictability. In other words, they automate cognitive tasks within the framework of ethical decisions previously reserved to humans. As a result, amid geographical scholarship there has been a conceptual shift from algorithms as invisible mediators of social life and spatial forms, to algorithms as autonomous agents whose decisions produce tangible socio-spatial effects. By using self-driving vehicles as case study and drawing on Hayles’ posthumanist epistemology, this paper aims to discuss further the agency of algorithms on two interrelated grounds. Firstly, it suggests a systemic and materialist perspective on Artificial Intelligence, as the automation of the driving task is enabled by a constellation of technologies (sensors, actuators, algorithms, mapping systems) concurrently operating at multiple spatiotemporalities. Secondly, algorithms are theorised as machinic ‘cognizers’ embedded within cognitive assemblages, wherein agency is distributed and neither algorithms nor humans can be said to operate within fully autonomous realms. As the discussion of self-driving vehicles attempts to show, automated decisions result from complex human-machine interactions and thus are nearly impossible to attribute to a single sovereign authorship.
Facial Recognition and the Automation of Security/Marketing Assemblages: Geographies of Anxiety and Pleasure in Music Festivals
Harrison Smith (Newcastle University, UK)
Jeremy Crampton (Newcastle University, UK)
Kara C. Hoover (University of Alaska, Fairbanks, USA)
J. Colette Berbesque (Roehampton University, UK)
This paper draws on preliminary survey results, media discourses, and industry materials around automated facial recognition for large scale events to theorize the affective geographies of surveillance and artificial intelligence. We focus specifically on unpacking affective relations of anxiety and pleasure in music festivals. Such events are typically seen as risky or unpoliced spaces for conspicuous consumption free from normative constraints of everyday life. Recently, however, these spaces have been problematized along several axioms of risk and security, and have made festival patrons, organizers, and the general public more aware of potential safety concerns. In turn, this has created a market for festival security technologies, particularly facial recognition and wearables that seek to automate festival securitization. However, these technologies have also been promoted concurrently for marketing applications to enhance festival atmospheres for brands and social media. We explore these technological, economic, and affective politics by presenting findings from festival participants concerning their attitudes and behaviours towards automated surveillance and theorize facial recognition around affective geographies of anxiety and pleasure.