Call for papers: “Making Machine Learning Worlds”, Royal Geographical Society conference 2023, deadline for abstracts 10 March.
Call for papers: Making Machine Learning Worlds
Royal Geographical Society with IBG Annual International Conference, London 30th August – 1st September 2023.
Convenors: Louise Amoore and Ludovico Rella (Durham University)
This session interrogates the contemporary geographies of machine learning models through the lenses of worlding and world-making. It seeks to understand how algorithmic models are becoming ubiquitous tools to navigate through our world, but yet also generating new and different worlds. When synthetic data such as “non-existing” faces are used to train facial recognition algorithms for policing, or when large language models are used to generate texts about anything including legal advice, it is clear that these models are changing the spaces, the means, and the outcomes of material instantiations of society, law, and political economy (Delaney 2010). When reinforcement learning and machine learning algorithms are used to reinvigorate agent-based modelling, large simulations, and digital twins – in turn used to model anything from the human brain to climate change to pandemic spread and impact, to the metabolism of smart cities – AI agents’ navigation of digital spaces and the lived experience in “real space” become so intertwined that machine learning can be said to play a substantial part in making and remaking worlds. At the basis of this process, a concentration of the “inter-referencing” (Roy 2011, 10) of machine learning models around a few core “foundational models” also poses questions around who gets to decide which worlds get to be generated, and which ones are foreclosed.
If “mediation is at the core of practices of world-making that engender shared worlds of lived experience” (Meyer 2008, 5), then what is the ethico-politics of algorithmically-generated worlds of shared lived experience? The agential power that algorithms exert on the world resonates with ideas of worlding and world-making: they represent “worldviews”, “not a view of the world but the world understood as a view” (Heidegger [1951] 1976, 350 in McCann et al, 2013, 585). Understood as “giving accounts of themselves” (Amoore, 2020: 18) the algorithms of reinforcement learning, for example, resonate with urban geographical research on worlding as “the ways that cities [or algorithms] assert their local economy and culture as positioned within global flows of capital, people and information” (Burns et al. 2021, 464; Simone 2001). Somewhat more prosaically, the world that machine learning algorithms inhabit also requires material making, i.e., the construction of ever more capillary forms of infrastructures spanning from the Cloud to the Edge, the multiplication of sensors, and specific forms of hardware architectures to accelerate extremely intensive forms of computation. This material world-making, in turn, has serious environmental and social consequences.
The session seeks papers exploring topics including, but not limited to, the following:
• Geographies of generative language and image models
• Theory and concepts of generativity and emergent effects.
• Large scale simulations and digital twins
• Geographies of model application domains, including:
• Law and Policing
• Smart Urbanism
• Climate Change adaptation
• Economic Forecasting
• Logistics and Self Driving Vehicles
• Hardware and infrastructures for machine learning
• Relationship between machine learning, simulation, and gaming worlds
We are eager to create a diverse and inclusive session. Submissions from early career scholars, research postgraduates, and those with underrepresented voices in the ‘digital geography’ debates, are particularly encouraged.
Please send your paper title and an abstract of no more than 250 words to louise.amoore@durham.ac.uk and Ludovico.Rella@durham.ac.uk by 10 March 2023.
References
Amoore, Louise. 2020. Cloud Ethics: Algorithms and the Attributes of Ourselves and Others. Durham: Duke University Press.
Burns, Ryan, Victoria Fast, Anthony Levenda, and Byron Miller. 2021. ‘Smart Cities: Between Worlding and Provincialising’. Urban Studies 58 (3): 461–70. https://doi.org/10.1177/0042098020975982.
Delaney, David. 2010. The Spatial, the Legal and the Pragmatics of World-Making: Nomospheric Investigations. London: Routledge-Cavendish. https://doi.org/10.4324/9780203849101.
Heidegger, Martin. (1951) 1976. ‘The Age of the World View’. Translated by Marjorie Grene. Boundary 2 4 (2): 340. https://doi.org/10.2307/302139.
McCann, Eugene, Ananya Roy, and Kevin Ward. 2013. ‘Assembling/Worlding Cities’. Urban Geography 34 (5): 581–89. https://doi.org/10.1080/02723638.2013.793905.
Meyer, Birgit. 2008. ‘Materializing Religion’. Material Religion 4 (July): 227–227. https://doi.org/10.2752/175183408X328325.
Roy, Ananya. 2011. ‘Urbanisms, Worlding Practices and the Theory of Planning’. Planning Theory 10 (1): 6–15. https://doi.org/10.1177/1473095210386065.
Short, Kathy G. 2012. ‘Story as World Making’. Language Arts 90 (1): 9–17.
Simone, AbdouMaliq. 2001. ‘On the Worlding of African Cities’. African Studies Review 44 (2): 15–41. https://doi.org/10.2307/525573.
Stewart, Kathleen. 2020. ‘Afterword: Worlding Refrains’. In The Affect Theory Reader, edited by Melissa Gregg and Gregory J. Seigworth, 339–54. Duke University Press. https://doi.org/10.1515/9780822393047-017.