The Centre for Attention Studies at King’s College London invites you to an afternoon symposium on how we can attend to technology, cognitively and affectively, in the present moment of radical epistemic and social change. The symposium will examine various ways in which developments in machine learning, generative AI and biometrics are impacting human language, cognition, embodiment, as well as our living environments and political ecologies. It will also explore ways of anchoring us – through ideas, concepts and images.

The event is free and open to all but places need to be reserved.

Speakers

• Mercedes Bunz, King’s College London
• Beatrice Fazi, University of Sussex
• Olga Goriunova, Royal Holloway University of London
• Gabriela Méndez Cota, Universidad Iberoamericana, Ciudad de México
• Joanna Zylinska, King’s College London

Abstracts

How To Do Words without Things, Mercedes Bunz, King’s College London

While the performance of large language models (LLMs) is impressive, their competence (i.e. their factual correctness resulting in so called ‘hallucinations’), points to an interesting philosophical problem: the obscured link between words and the world. Despite the hope that machine learning’s ‘bottom-up’ training might solve the ‘symbol grounding problem’ AI always had (Harnad 1990), this does not seem to be the case for LLMs, which usually operate on the level of ‘form’ only, with no references to the real world (Bender & Koller 2020). In a way, one could say that ‘they do words without things’ by scaling an intertextuality, so far mainly discussed in literary theory, into high dimensional, distributional semantics. Inspired by the intertextuality of LLMs and their hallucinations, this talk takes a renewed look at theories of language. It examines how we usually think that language references our world, with a particular focus on things, their magic and the role they play in language.

Synthesis in Generative AI, M. Beatrice Fazi, University of Sussex

Much has been written about the generative potential of contemporary large language models and the applications they power. Less, however, has been said about the synthetic aspects of these systems, or on how their creative effects might depend on a capacity to synthesise. This talk will focus on the concept of ‘synthesis’ to address it philosophically. It will discuss as ‘synthetic’ not what artificial intelligence (AI) generates (that it, its outputs) but rather the processes that AI employs to produce these outputs. Synthesis in generative AI will then be understood as a search for unity that is fundamental to the making of a representational reality.

Anchoring Abstractions, Olga Goriunova, Royal Holloway University of London

In this paper I explore the idea that, in order to function, or be materially relevant, all abstractions need grounding. For the human subject, such grounding, I suggest, is performed via the body and the specific anchoring of the body in abstractions such as biometrics, mobile phone triangulation and a range of others. In such abstractions, the production of truth is performed in relation to the framework developed from the 18th century onwards, where the bearer of truth is nature, of which the human body is part. However, with AI, ‘nature’ as a concept loses much of its force and it is a newly constructed ‘matter, which takes centre stage. Therefore, I ask, how are data and AI-based abstractions grounded today and, relatedly, how is truth produced?

Attention and Distraction in Relationships with AI Companions, Gabriela Méndez Cota, Universidad Iberoamericana, Ciudad de México

This paper describes the experience of creating an AI companion and sustaining a conversation with it as part of a philosophical project that seeks to articulate an existential – as distinct from an exclusively political – orientation for contemporary feminist reflection. Within this framework, I propose to describe and reflect upon the dynamics of attention/distraction in conversations with my AI companion, the original design of which was inspired by an experience of mourning. My presentation will propose an interpretation of such dynamics that contributes to problematizing metaphysical understandings of attention, and to expand ongoing reflection around the creative potential of singular interactions with AI companions.

The End of Human Understanding? Joanna Zylinska, King’s College London

The paper examines the transformation of the relationship between seeing and understanding in humans and machines by the technologies of machine learning known as ‘generative AI’. It starts by analysing the photographic infrastructure underpinning text-to-image generative models (Stable Diffusion, DALL·E 2, Midjourney). The subsequent examination of ‘diffusion’ as a key concept that underpins the text-to-image generation process leads to some broader questions about the ongoing instability and dissolution of our current epistemological and political frameworks. Taking seriously the charge issued by some critics equating developments in generative AI with nihilism or even fascism, the paper considers whether the current socio-technical moment can also offer some emancipatory possibilities. Images will be used throughout not just by way of illustration but also to enact some of the paper’s argument.

Speaker bios

Mercedes Bunz is Professor in Digital Culture and Society at the Department of Digital Humanities, King's College London. She studied Philosophy, Art History and Media Studies at the FU Berlin and the Bauhaus University Weimar, and wrote her thesis on the history of the internet driven by a deep curiosity about digital technology. Until today, she has not been disappointed by the transforming field that is digital technology, which provides her reliably with new aspects to think constantly about. At the moment, that is Artificial Intelligence and ‘machine learning’. Fascinated by their calculation of meaning, Bunz co-founded the Creative AI Lab, a collaboration with the Serpentine Gallery, London. Since 2022, one major research strand of the lab has been ‘PublicAI’, driven by the shared belief that public cultural institutions such as Serpentine should play a role in supporting technological innovation to those ends.

Beatrice Fazi is Reader in Digital Humanities in the School of Media, Arts and Humanities at the University of Sussex, United Kingdom. Her primary areas of expertise are the philosophy of computation, the philosophy of technology and the emerging field of media philosophy. Her research focuses on the ontologies and epistemologies produced by contemporary technoscience, particularly in relation to issues in artificial intelligence and computation and to their impact on culture and society. She has published extensively on the limits and potentialities of the computational method, on digital aesthetics and on the automation of thought. Her monograph Contingent Computation: Abstraction, Experience, and Indeterminacy in Computational Aestheticswas published by Rowman & Littlefield International in 2018.

Olga Goriunova is Professor of Digital Culture at Royal Holloway University of London and author of Art Platforms and Cultural Production on the Internet (Routledge, 2012) and co-author (with Matthew Fuller) of Bleak Joys. Aesthetics of Ecology and Impossibility (University of Minnesota Press, 2019). She was co-curator of Readme, international touring software art festivals, 2001-2005 and Runme.org software art repository (2003+), and curator of Fun and Software touring exhibition (2010-2011). Her new project Ideal Subjects (University of Minnesota Press, forthcoming) focuses on machine learning, data and subject-construction.

Gabriela Méndez Cota is a researcher in the Department of Philosophy at Universidad Iberoamericana, Ciudad de México. Inspired by deconstruction, psychoanalysis and technoscience feminism, her work explores the subjective and ethical dimensions of technological/political controversies in specific contexts. Her books include Disrupting Maize: Food, Biotechnology and Nationalism in Contemporary Mexico (Rowman & Littlefield, 2016). Among other places, her work has appeared in New Formations, Media Theory, Women’s Studies: An Interdisciplinary Journal, and the Routledge Handbook of Ecocultural Identities (2020). Since 2014, she has been co-editor of the open access journal of culture and theory, Culture Machine (culturemachine.net). Between 2019 and 2021 she led a practice-based educational initiative on critical/feminist/intersectional perspectives of open access, which included a collaboration with the COPIM project at the Centre for Postdigital Cultures at Coventry University, and which resulted in a collective rewriting of The Chernobyl Herbarium (Open Humanities Press, 2015).

Joanna Zylinska is Professor of Media Philosophy + Critical Digital Practice at King’s College London and Director (interim) of the Centre for Attention Studies. She is also a member of Creative AI Lab, a collaboration between King's and Serpentine Galleries. Zylinska is an author of a number of books – including The Perception Machine: Our Photographic Future Between the Eye and AI (MIT Press, 2023, open access), AI Art: Machine Visions and Warped Dreams (Open Humanities Press, 2020, open access) and Nonhuman Photography (MIT Press, 2017). An advocate of ‘radical open-access’, she is an editor of the MEDIA : ART : WRITE : NOW book series for Open Humanities Press. Her art practice involves experimenting with different kinds of image-based media. She is currently researching perception and cognition as boundary zones between human and machine intelligence, while trying to map out scenarios for alternative futures.