In the week of March 20th 2023 myself and a shifting ensemble of colleagues and students, Dr Dylan Yamada-Rice, Jing Wang Thomas, James Edward Marks, Andy Cheung and Jiahe Lin, took part in an online performance called ‘Monstrous’, as part of the Mozilla Festival (MozFest), 2023. Throughout the festival dozens of participants came to the five performances, joining us in singing, discussion and virtual dancing online. Monstrous was loosely centred around Herman Melville’s novel Moby Dick (1851), but it also incorporated ideas from C.L. R. James’s book, Mariners, Renegades and Castaways: The Story of Herman Melville and the World We Live in (1953). James’s book draws out comparisons between the factory whaling ship run by Captain Ahab, and the exploitative extraction of racialised industrial capitalism.
One key question in centring Moby Dick at MozFest, was, how can we surface analogies between 19th Century whaling practices and contemporary (late) capitalism, particularly, the hype cycles, working conditions and trajectories of AI? Logically implicated in that goal was the question of how best to engage people with these issues within a virtual environment. Artaud’s ideas about ‘the abolition of the auditorium and the stage to create a single playing space with no barriers between audience and performers’ (Artaud, 1978, Tripney, 2017), as well as Brechtian Alienation, or A-effect (Brecht, 2018) were useful frames of reference in forming a methodology and specific strategies. The aim of Brecht’s A-effect, also known as the Verfremdung or V-effect, was ‘to make the spectators adopt an attitude of inquiry and criticism in their approach to the incident. The means were artistic. The first condition for the V-effect’s application to this end is that stage and auditorium must be purged of everything ‘magical’ and that no ‘hypnotic tensions’ should be set up. This ruled out any attempt to make the stage convey the atmosphere of a particular place (a room at evening, a road in the autumn)’ (Brecht, 212, 2018).
But these methods are not without important critiques and contradictions, productive tensions which we worked with throughout the festival. Even Brecht himself confusingly alludes to the way Chinese performers use magic to create a V-effect(Brecht, 181, 2018). Rancière argued that audience reception is inherently more productive than Brecht’s conception of it. As Alston reminds us, for Rancière, ‘what these two paradigmatic conceptions of productivity overlook is the inherent productivity of reception. Audiences watch, listen, decode, cogitate, imagine, feel, hope and desire. These are all productive things to do in the sense that these actions produce meanings among a gathered audience of individual spectators, each ‘refashioning’ performance in their own way'(Alston, 18, 2016). Alston writes of Brecht’s critical approach to theatre, though which he ‘designed and mobilised dramaturgic and aesthetic strategies to awaken this kind of audience productivity’, while Artaud ‘proposed ideas for staging proto-immersive theatre, as these ideas might be understood today, so as to ‘cruelly’ jolt audiences out of docility’ (Alston, 18). For Artaud, Alston states, ‘an important aim was to facilitate the audience’s realisation that fiction is not what they encounter within the ritual of theatre, but in the socio-culturally coded world outside of the theatre’ (Artaud, 1958, pp. 96–7). Between immersion and alienation, magic and realism we formed our own approach to both engaging audiences and creating spaces for discourse with them. Artaud and Brecht provided usefully tense (and in tension) approaches to critically engaging audiences, within and beyond immersive entertainment.
Monstrous at MozFest
MozFest is advertised as an activist event to address ‘Trustworthy AI’. The question of trust – about what we are told by the media and what exactly is unfolding in machine learnt processes – is an urgent one. During the period of planning and performing Monstrous, new versions of Chat GPT were released, and with each new release came more media hyperbole, as well as more fear within Higher Education and wider society about the trajectory of large language models, as urgently addressed by Bender et al (2021). A range of writers reported on the oppressive working conditions of Kenyan and other workers who pre-processed Chat GPT data and wider labour conditions for those processing machine learning models (Chandran et al, 2023). Commentators also explored the flattening effect of AI generated imagery and its colonial systems of knowledge (Jenka, 2023, Sojit Pejcha, 2023).
These writers and my own work with Machine Learning solidified the intention of creating a contingent space for participants to discuss our understandings and questions about AI. I also researched and implemented a range of approaches to performing, using AI generated shorts plays and sea shanties, while Yamada-Rice recognised the need to simplify some forms of our online interaction, inviting audiences to clap or tap tables and other surfaces. As the planning stage progressed, I wondered if a revised approach to Anti-Theatre was relevant to the potentially reductionist processes of AI; could some of the strategies of Anti-Theatre help us to generate different types of discourse about Machine Learning technology? Hayman writes: ‘The anti-play is less mimetic than satirical, not so much a story about life in particular place at a particular time as an object in its own right, non-referential, implicitly denying the feasibility of referential art’ (Hayman, xii, 1979). Questioning the stability of knowledge represented by AI, such as hard and fast categories of emotion, gender, age and ethnicity is entangled with wider questions of stable representation. Questions of power and privilege are of course, also always implicated in these issues. Many critics of AI point to its function as that of reinforcing the status quo (MacQuillan, 2022, Noble, 2018, Benjamin, 2019). In 1979, Hayman wrote of Anti-Theatre as arising from an ‘aversion to current practices in literature, art and theatre’ which ‘inspired an alternative method of procedure’ (Hayman, xi, 1979).
While Monstrous deployed some immersive features, such as music and a realistic skybox and circus tents, at times it was also often non-immersive, arguably eschewing mainstream conceptions of a futuristic photo-realist Metaverse. To achieve this movement between alienation and illusion we deployed a range of strategies, including:
- emotional distancing, detachment from characters via non emotive acting
- critical reflection on social and political issues
- banners, signs and questions
- addressing the audience directly
We also used the following approaches:
- drawing audience attention to the mediated nature of the Hubs experience, such as its interfaces and upload processes, bandwidth fluctuations and sonic feedback
- Use of unrealistic and exaggerated characters (generated by Machine Learning), such as the two odd models of Moby Dick
- AI generated ‘plays’ with draft versions of scripts and prompts visible to audiences, who were invited to read through them as impromptu performers
- the use of fragmented, non-linear narratives that do not mirror conventional storytelling structures or expectations
These methods were intended to remind audiences that they were watching a performance, to confound expectations of Metaverse immersion and focus instead upon our questions and thoughts about AI. Connors (2019), pinpoints the radical nature of an apparently anti-entertainment stance for directors and actors whose: ‘audiences live in a world drenched in entertainment and celebrity dazzle. This world of mass distractions has almost entirely absorbed politics and even, in some areas, science’ (2019). The sometimes non-immersive design of the Monstrous virtual spaces as well as the V-effect tactics were influenced by my teaching and previous research. Brecht’s goals resonate with many of my own pedagogic goals, although they conflict with other intentions, such as creating experiences which funders and collaborators desire, to create spectacles and amusements and my own interest in creating magical experiences, which chimes much more with Artaud’s Anti-Theatre.
My experience of teaching interaction design and VR storytelling in the context of UK Higher Education, is that students often seem to place great emphasis on a construct of immersion for its own sake, without articulating how and why it matters, or what its downsides might be. To interrogate this with them, I often ask students in what way they experience digital immersive virtual spaces, do they make them feel different from, say 2D spaces, or analogue books? One student once commented that the scale of a set I developed in Mozilla Hubs for Brecht’s Life of Galileo had made them feel small in the universe; it seemed to be a positive comment, as if the space had moved them in a way which resonated with Brecht’s characters first looking through a telescope. My understanding is that Brecht wanted his audiences to remain detached from the characters on stage, a point he made with his first ever production in 1922, of the Life of Galileo (Brecht, 2014). Brecht hung banners in the theatre instructing the audience not to become too emotionally involved. He wanted to engage his audiences intellectually, in the social and political ideas of the play, not lull them with mere entertainment and escapism.
In teaching digital storytelling, arts creativity and games development, I sometimes ask students if virtual spaces can support a shift in the order of power between, for example, students and lecturers, and if so, what other kind of power relations might emerge? We have discussed the fear expressed in Brecht’s Life of Galileo, that peasants, exposed to the enhanced vision of telescopes, might start to doubt the given order. In Monstrous I wanted to engage audiences similarly, in exploring our ideas about AI and power. But with AI the issue of reductionist systems of knowledge is significant, Brecht is arguably part of the same logical positivist tradition (Giles, 1995), with a belief in scientific vision and empirical knowledge which predates a postmodern crisis of confidence in singular, objective truth.
For Artaud the goal of theatre was to provide a heightened, non-rational, non conditioned experience that might jolt audience assumptions about reality. Magic was a key part of achieving this goal, and the theatre a locus for new types of ritual and alternative realties to emerge. This assault upon one system or disembodied regime of rational, disembodied knowledge chimes with Machon, for whom:
‘A vital component of immersive theatre, then, is the fact that it revels in an energised liveness and the consequent live(d)ness of the performance moment. To clarify, whatever forms the imaginative journey through the event takes, the sensual worlds created for and through immersive practice exploit the unique power and theatricality of live performance as an artistic medium. A very real exchange of energy between humans exists within the immediacy of the live and ongoing present of performance. A sense of being alive – more significantly, an awareness of the lived moment – is keenly felt. This connects with Gilles Deleuze’s theories of “immanence” as “pure presence” (Deleuze, Francis Bacon 52) brought about by visceral artworks’ (Machon, 2016).
Such immanence and embodied experience is quite different from the logic of AI and even from Brecht’s analytical, arguably positivist intentions. Brecht wrote: ‘How can the attitude of the opera audience be characterized, and can that attitude change? Rushing out of the underground station, eager to become like putty in the hands of the magicians, grown men who have proved themselves ruthless in the struggle for existence rush to the theatre box-offices. As they hand in their hats at the cloakroom’ (Brecht, 76, 2018).
Within the Hubs space we used many conventionally immersive assets, such as a 360 video sphere depicting a roller coaster ride (a metaphor for Block Chain’s inefficiency). Music and spatialised sound effects also constructed a realist approach to space. But we also deployed magic (in the sense of optical illusions and sleight of hand) in the run up to MozFest on social media, as adverts for the performances but also as provocations, to draw out analogies between AI, advertising and corporate hype.
The performance space for Monstrous consisted of two virtual circus tents, one of which contained an oddly distorted AI generated whale and play scripts generated by Chat GPT. The other tent contained images of historical artificial intelligences, such as the Mechanical Turk and the Digesting Duck, visually re-generated by text to image machine learning algorithms. The circus tents were located in a larger virtual amusement park designed and developed by Dare as part of our MozFest space, called the AI-Musement Park. AI-Musement Park acted as a hub to connect a range of other events and performances, as well as a portal to the main MozFest home space.
Throughout each performance we were supported by MozFest volunteers, these volunteers, present as avatars, welcomed audiences and guided them towards the starting tent. They also joined in singing and dancing, kickstarting audience participation and instigating conversation about the theme of AI and extraction. Feedback was solicited by volunteers via a formal feedback mechanism. We are still awaiting a detailed breakdown of these responses, but ad hoc comments were positive, stating that the performances had spurred new lines of thought about AI, especially its environmental impact and labour conditions. Many audience members found the sounds we used for navigating the space intrusive, Dare was constantly trying to adjust the sound, to provide accessible navigational clues but also to evoke shifts in register. Live instruments were played as audiences arrived and left, such as a toy accordion, electric guitar and clarinet. One person seemed confused by the sound of the accordion, ‘how did you do that ?’ they asked Dare, perhaps expecting a highly technical answer, ‘I’m just playing an accordion in front of my laptop’ Dare answered. Anecdotally, it seems analogue approaches are more astonishing to audiences then the many types of technological mediation we are now accustomed to online.
Our intention is not to idealise the type of audience participation encouraged by Monstrous and the strategies described here, we are aware of critiques of such approaches, for example, Schneider, who has critiqued immersive theatre from a feminist perspective. arguing that immersive theatre often reinforces gendered power dynamics, objectifying the bodies of performers and audience members. (Schneider, 2011). But we do think (and hope) stratgeies arising from Anti-Theatre, Theatre of Cruelty, magic and Alienation, in productive tension, can be useful for other researchers and for ourselves in future work.
Many thanks to
Dylan Yamada-Rice, Jiahe Lin, Jing Wang Thomas, Ariel Cao, Andy Cheung, Marius Matesan, James Edward Marks and Ken Goffman (RUSirius).I would like to thank all the MozFest volunteers, especially Livvy and Michael Morran, Temi Popo and Tyler Munyua,
Monstrous was based inside the AI-Musement Park, a Mozilla Hubs virtual social space from which visitors could enter three other spaces: Fungi See Sea Fun by Jiahe Lin, Copenhagen Reimagined, 1940, by Jing Wang Thomas, and Mondo 2000, RU Against NFTS, by RU Sirius (Ken Goffman), James Edward Marks and Marius Matesan. Ariel Cao ran zoom based sessions on Girls /(LGBTQ+) in STEAM. Tyler Munyua presented a painting and link to their Obsidian Circus Gallery.
References
Alston, A. (2016) Beyond Immersive Theatre: Aesthetics, Politics and Productive Participation, London: Palgrave Macmillan.
Artaud, A. (1978) The Theatre and its Double, Croydon: Éditions Gallimard.
Bender, E. M., Gebru, T., McMillan-Major, A., Shmitchell, S. (2021) ‘On the dangers of stochastic parrots: Can language models be too big?’ In Proceedings of the ACM Conference on Fairness, Accountability, and Transparency. 610–623.
Benjamin, R. (2019) Race After Technology, Cambridge: Polity Press.
Brecht, B. (2018) Brecht on Theatre: The Development of an Aesthetic by Bertolt Brecht, London: Bloomsbury.
Brecht, B. (2014) ‘Life of Galileo’. In Collected Plays: Five, translated by John Willett. London: Bloomsbury Academic.
Chandran, R., Smith, A., Ramos, M. (2023) ‘AI boom is dream and nightmare for workers in Global South’, Context, https://www.context.news/ai/ai-boom-is-dream-and-nightmare-for-workers-in-global-south?utm_source=pocket-newtab-global-en-GB&utm_source=twitter&utm_medium=social&utm_campaign=context-share
Connors, J. (2019) ‘The Cleveland Play House makes its debut at the Allen Theatre with a splendid production of ‘The Life of Galileo’, Cleveland.com https://www.cleveland.com/arts/2011/09/the cleveland play house makes.html
Hayman, R. (1979) Theatre and Anti Theatre, London: Secker & Warburg Limited.
C.L.R. James (1953) Mariners, Renegades and Castaways: The Story of Herman Melville and the World We Live in.
Steve Giles. (1995) ‘Bertolt Brecht, Logical Empiricism, and Social Behaviourism’, The Modern Language Review, 90(1), 83–93.
Jenka (2023) ‘AI and the American Smile
How AI misrepresents culture through a facial expression’. Medium.com https://medium.com/@socialcreature/ai-and-the-american-smile-76d23a0fbfaf
Herman Melville (1851) Moby Dick, London: Richard Bentley.
Sojit Pejcha, C. (2023) ‘The work of art in the age of algorithmic optimization’, Document Journal, https://www.documentjournal.com/2023/03/ai-image-generators-art-intellectual-property-copyright-stable-diffusion-dall-e-glaze-spawning/
MacQuillan, D. (2022) Resisting AI: An Anti-fascist Approach to Artificial Intelligence, Bristol University Press.
Machon, J. (2016) ‘Watching, Attending, Sense-making: Spectatorship in Immersive Theatres’. Journal of Contemporary Drama in English, vol. 4, no. 1, 2016, pp. 34-48. https://doi.org/10.1515/jcde-2016-0004
Noble, S., U. (2018) Algorithms of Oppression How Search Engines Reinforce Racism, New York: NYU Press.
Rancière, J. (2009) The Emancipated Spectator, London: Verso.
Schneider, R. (2011) Performing Remains: Art and War in Times of Theatrical Reenactment, London: Routledge.
Tripney, N. (2017) ‘Antonin Artaud and the Theatre of Cruelty’, British Library, https://www.bl.uk/20th-century-literature/articles/antonin-artaud-and-the-theatre-of-cruelty
short stories etc 