By Maisie Ridgway
Intelligent Futures was a postgraduate and ECR conference, supported by CHASE DTP and Sussex Humanities Lab. Over the course of two days, the conference challenged researchers to find original, philosophical and cultural approaches to Artificial Intelligence. The interdisciplinary explorations spanned the social sciences, informatics, psychology, art, literature and more, promoting critical and speculative engagements with technical cognition.
Thomas Nyckel of the Technical University Braunschweig led the first panel, which sought to engage with AI from a philosophical standpoint. Nyckel’s paper fostered an alternative approach for understanding the processes of digital devices and computation through the variable definitions of the ‘rule of thumb’. Nyckel’s identified two categorisations of the rule of thumb – Frederick Taylor’s conception, concerned with exact scientific results, and Alan Turing’s interest in approximate computational methodologies. What materialised was a synthesis of ideas that challenged the nature and myth of scientific exactitude and complicated the binary of workman and machine, or the approximate rule of thumb and exact scientific methods. Mattia Paganelli followed Nyckel with a paper on the misnomer of ‘artificial’ when speaking of non‑human intelligence. Paganelli argued that the term shored up the binary between subject and object by attributing an a priori definition of the perimeter of possibility for a given system. Rather than artificial, Paganelli suggested the term ‘thinking machines’, understanding intelligence as a process of learning, plasticity and openness to name a few key attributes.
The next panel united speakers under the common theme of ethics. Camilla Elphick of the University of Sussex spoke about her ongoing project to develop an AI chatbot, named Spot, as a means by which victims of work place sexual harassment can report their experiences. Spot’s distinctly unhuman style of engagement meant that victims did not feel embarrassed, judged, scrutinised or pressured, garnering more accurate information. On an entirely different note, Marek Iwaniak’s paper explored how theologies and pre-technological religious imaginaries could engage with the novel challenge of AI. Iwaniak speculated on possible intersections between various religions and AI such as a Buddhist approach to technical cognition whereby the ultimate aim of developers could be an AI consciousness of pure bliss.
The final panel of the day explored changing ideas of writing, asking whether an anthropocentric conception of literary creativity obscured other forms of nonhuman creative generativity from view. John Phelan of the Open University investigated if AI could ever critically appreciate poetry or poetic significance outside of large sample readings of rhyme schemes for example. Emma Newport, from the University of Sussex, considered digitised end-of-life writing via the unusual case of the popular Mumsnet contributor IamtheZombie, whose posthumous in memoriam comments from fellow Mumsnet users formed an innovative kind of obituary, a cellular tissue of text that democratised the death process and made end-of-life writing a collaborative act.
Joanna Zylinska’s keynote speech, entitled ‘Creative Computers, Art Robots and AI Dreams,’ drew together the prevalent themes of the day, excavating the myth of the robot to determine that humans, especially the great artists amongst us, have always been technological. Zylinska used various art-AI projects, including Taryn Southern’s AI generated music and The Rembrandt Microsoft Project (involving AI creating an imitation Rembrandt painting), in order to interrogate the value of AI artistic production, delineate how our senses construct the world we inhabit, and ask what it means that seeing, for example, no longer requires a human looker. The result was a critique of AI as that which exponentially amplifies our desires and therefore marketability, as well as an optimism for the possible joys of robotic and AI art, described by Zylinska as ludic creation.
Day two began with a panel on epistemology. Emma Stamm of Virginia Tech presented a paper on the renaissance of psychedelic science and the implications this field could have for AI. Stamm described the importance of qualitative over quantitative methods of research, arguing that qualitative research into psychedelic drugs problematises the positivist and generalising principles of machine learning as a basis for AI. Juljan Krause from the University of Southampton followed with his thoughts on representations of quantum computing in popular science discourse. Krause offered an interesting overview of how the emerging technology of quantum computing functions differently from present modes of computation. He then explored media representations of quantum computation, which portray an ongoing quantum computational arms race (led by China).
The penultimate panel revolved around the theme of aesthetics. In his paper on art and artificial intelligence Michael Haworth reconfigured the relationship between human and machine as an equal coupling, a functional interdependence manifest in the example of AARON, the painting and drawing robot. Haworth relayed how the creativity of AARON rests in the relation between the program and programmer, an interplay that performs a structural shift from the human as tool user to the human as engineer and organiser. Following Haworth, Dominique Baron-Bonarjee presented a performative lecture during which she enacted Liquidity, an embodied, meditative practice that she designed to explore ideas of free time in an increasingly automated age. Baron-Bonarjee monitored the activity of her brain throughout the lecture using a MUSE headband which measured her brain waves.
Memory and Time formed the focus of the final panel, offering an eclectic range of thoughts on automation and redundancy, logistics and inscribed narrative time. Kieran Brayford discussed the possible ramifications of mass technological unemployment and forced leisure time, coming to similar conclusions as Phelan with regards to the limitations of automated agents as unable to explain significance. Eva-Maria Nyckel’s followed Brayford’s more general overview with a specific example of automated industry in the form of Amazon’s anticipatory shipping method. Nyckel used a patent for the shipping method as the basis to unpack the potentially huge impact it could have on the temporality of logistical processes. In a change of topic, Daniel Barrow of Birkbeck University turned to contemporary experimental fiction and the technological, inhuman agency of Big Data, which finds its form in a static and autonomous narrative time.
To close the conference a number of notable Sussex based academics and researchers came together for a round table discussion. Taking part were Caroline Bassett, Peter Boxall, David M. Berry, Beatrice Fazi, Simon McGregor and Michael Jonik, all of whom were asked to choose one key word that would explain their research. Boxall chose the word ‘artificiality’ as a recourse to explore its supposed opposite- the human. He offered the idea of the Augustinian human as somewhere between beast and angel, a figure made legible according to taxonomies that we map onto it. McGregor’s word was ‘alien’. As a cognitive scientist McGregor explained that the world is full of minds that cannot be reduced to their materiality and are eternally separated by an unbroachable void. Any attempts to understand these minds would aid efforts to predict their behaviour but would also bring us further away from other humans. Other keywords included intelligence, contingency, mastery, and critical reason, with final thoughts settling on the infinitude of computation, the passive intellect of algorithmic infrastructure, and the potential for the world to continue in the absence of humans.
Over its two days, Intelligent Futures gathered a range of richly provocative critical interventions, and created spaces for stimulating discussion of Artificial Intelligence. It demonstrated the importance of further critical, interdisciplinary study of Artificial Intelligence, as it continues to inform, and to transform, the societies we live in.
Maisie Ridgway is a CHASE funded PhD student at the University of Sussex. Her research interests include ideas of pre-digital poetics from Joyce to present, the viral vitality of language and the points at which science and literature intersect, mutually informing each other.