Unit 3
The Research Festival : ‘S0_P0T3NT_4RT’
For the Research Festival, I’ve made a video titled ‘S0_P0T3NT_4RT’, which is a narrative structured with a tutorial format. The video is part tutorial on using Stable Diffusion AI video rendering and part tutorial on summoning and entrapping a demon. This video links my research into AI with occult themes and depicts my personal experiences of using AI whilst battling anxiety surrounding the potential dangers of emerging technologies. The comparison between these tutorials additionally invites absurd dark humour through the combination of the popularised tutorial format with the presentation of fictional occult information. As the video continues, the lines between the two rituals will begin to blur. The digital avatar of Ariel, who is giving the tutorial on Stable Diffusion video rendering is summoned by the Shakespearean, New England witch, styled loosely on a combination between myself and Prospero, who is giving instructions on the demon summoning. The demon familiar Ariel AI presents the witch with a choice: should she proceed and join forces with Ariel? Or drown her magical book and abandon utilising AI generation as artistic practice, given the moral complexities surrounding both contexts?
At this time in the research festival, I intend to offer my audience a choice as a live event after this initial screening. I will present to them two clay balls/pebbles each, to take as they walk into the screening: one white, one black. Based on an Ancient Greek system of pebble voting, I will present a bowl to the voters and ask that they drop a white pebble into it to vote ‘yes’ OR a black one to vote ‘no’ to the question: ‘should contemporary artists continue to use generative AI art within their practice?’ Depending on the viewer’s perspective, artists might seek to change the injustices of databases using images of people and artworks without consent, or artists might only aggravate this problem by indulging in these technologies through their practice. The video has two endings, though only one will be screened at the research festival. The ending selected and screened will be dependent on the vote that the viewers cast. The first ending for ‘yes’, sees an abstract collage that merges the witch with the spirit, her familiar, to create a united, cyberfeminist entity that draws together the binary principles of the user and the used into one. The second ending for ‘no’, depicts a debate between the witch and the AI model, which ends in their ultimate separation. Both endings offer hope and tension, ensuring that there is no ‘right’ ending in this ongoing debate, with so many factors at play.
‘S0_P0T3NT_4RT’ Research Festival Transcript Extract:
Upload your video by dragging it to the space on the left, to FEED ME DATA. *contrastingly quiet* Where the folder symbols are. Make sure the video name doesn’t have spaces. MACHINES DON’T LIKE SPACES. WE CANNOT RE-READ- *GLITCH* Use underscores instead. Copy the uploaded video link. Paste the link into ‘video_init_path:’ under Video Input. Now AI can read the original footage input. Change ‘extract_nth_frame:’ to ‘2’. The AI will generate every second frame to match the frame of your video, your artwork, your data. The AI tends not to enjoy any figure much over ‘700’. These figures determine the size of the generation.*ALTERNATING SIZES* 512 is low for image quality. Change it to around ‘712’. Run Load Settings.
Prospera: It’s time to make your sacrifice. You have to give in order to receive, so, are we ready? We must shed blood. *lightening* Just a little of our blood, hmm? This will bind us to the power of this spirit and make it our blood-bonded familiar. The spirit shall surrender its power to us, and render us all powerful.
“Generative AI art is vampirical, feasting on past generations of artwork even as it sucks the lifeblood from living artists” (Centre for Artistic Inquiry and Reporting, 2023)
Change the ‘batch_name:’ each time you do this process. Change the ‘seed_behaviour:’ to the ‘fixed’ option.
This allows each generated image to follow the composition of the one before it, allowing for frames that flow in movement when placed together. Run ‘Save & Display Settings.
Prospera: *wiping blood away, dishevelled* Now we must keep our progress. We decide how much to control. We must alter the strength of the spell and smother all resistance. *crushing the feather* We must ’Change the ‘strength:’ to the desired figure. *writing something* The strength determines how much control the spirit will have over our project, our art.
*GLITCH SAYING:*
HOW CAN YOU CONTROL THIS?
YOU DO NOT CONTROL THIS.
ONLINE SACRIFICE.
The closer to ‘1’ the strength is, the closer to you I remain. For my own spell, I have discovered that ‘0.5’ was far too removed *voices deteriorates into a technological voice before returning to a human voice*. 0.8 is too close to you simply making up the imaginary. You’re not doing any magical thing at all. But between 0.65 and 0.7, 0.65…
I am complete.
Run the rest of these settings. Make sure the ‘image_path:’ and the ‘mp4_path’ will save the frames in your google drive. Run these settings. The red of the prompts section is where you can write your orders. Please type in a prompt. What would you have me be?
Prospera: What’s it to be? What will we make? it isn’t too late. *Turning to the camera* I need help to decide. It exists now… but we don’t have to use this. Then again, if we don’t, are we neglecting something? It’s real and it’s here. Maybe… it can be for all of us. What ought we to do?
Talos and The Sea Storm
The video above depicts a draft video work I created in preparation for the Research Festival. This work gathers together my research, including my research on John Ruskin’s plague cloud, on data and on Talos from Greek mythology. This range of sources is placed together under the overarching context of our relationship with data harvesting and emerging AI technologies. The work also introduces the character of the New England witch, who has gone on to play a major role as the main character in my final work for the research festival. A tutorial with the Camberwell tutor and artist, Jennet Thomas helped me to recognise how this character might help an audience to engage with the information my work for the research festival presents – humanising my work when learning about mystifying technologies and broad statistics. When reflecting on this work, I realised viewers may become invested when presented with a loose narrative concerning my research and reflection as opposed to an assemblage of my research. Narrative has always been key to my practice, which is why it seems important to include it in the research festival. Part of my progression over this unit has manifested itself in the discovery of practice as research. Through communicating my research artistically, I hope to encompass artistic, literary and technological contexts in a work that additionally appeals to the artistic community I’m a part of.
Talos and The Sea Storm Transcript:
‘John Ruskin gave a series of lectures called ‘The Storm Cloud of the Nineteenth Century” (Bridle, 2018, pp. 17) in 1884, at the London Institution. In these lectures, he presented his studies of the English sky, and his conclusion that an unprecedented wind was sweeping over England, bringing a new type of dark cloud with it. He called it a plague cloud.
Not rain-cloud, as he had described, but “a dry black veil, which no ray of sunshine can pierce.” (Bridle, 2018, pp. 18.)
Ruskin appeared to be concerned that this cloud had seemingly emerged from battlefields and sites of ‘societal unrest’, (Bridle, 2018, pp. 18.) though he had also acknowledged the increasing air pollution caused by industrialisation within the landscape of his observations. Still, to Ruskin, this cloud was potentially amoral… something perhaps, that could even be a representative of evil.
Something wicked this way has come.
William Shakespeare’s play The Tempest opens with a storm. Prospero, the sorcerer, ordered Ariel, an airy spirit, to create a sea storm.
The storm results in Prospero’s enemies from the court of Milan being shipwrecked on the shore of the island where Prospero had been living for twelve years, since the courtiers banished him.
Prospero was a Duke of Milan in Shakespeare’s play, before he was banished. Rapt in his studies as a user of magic, he neglected his state as he himself confesses in the play, allowing his brother to seize power and usurp him. (Shakespeare, 1999, 1.2.68.)
I am one such – partaking in my own secret studies and liberal arts. I have also neglected a measure of worldly matters and exploration, instead confining myself to the acts of researching and creating – fine art itself being magic to delve into. This is not, however, a narrative that is solely about art and magic.
In Greek Mythology, Talos was the first robot on earth. A bronze giant, sometimes winged, He was stationed to protect Crete’s shores, to wreck enemy ships by throwing rocks torn from a cliff face at them. He could also heat his body and perform a lethal embrace – burning alive those caught in his arms, should they come close enough. (Mayor, 2018, pp. 7)
This is not just a story of art and magic. This is a tale of artificial minds and bodies.
We gather on our ships on the ocean of the online, unaware, or uncaring of the spirit that cackles as it summons the electricity once belonging to Jove, or Zeus. We gather on our ships on the shoreline of a thousand data banks, waiting for the moment that Talos’ bronze, unfeeling eyes scans the digital map, registering our current location. Registering who we are, and what our histories were.
AI bears witness to our text. And while I write, the bits of data, 8 bits in a single byte, are multiplied, like the cells of an organism, ever dividing, ever inflating. In 2020, it is said that we generated as many bits on our planet as stars in our observable universe. (Butler, 2016, pp. 398 – 401.)
By 2170, if we could even venture to survive a thousand years, we would have collected enough bits to match the sum of every atom existing on the Earth. (Chadwick, 2020) A world inside a world. A universe of information. Perhaps _our_ universe belongs to _it_. An expanding plague. Our digital databank of dry, obsidian clouds.
AI and Animism
My research in Unit 2 flourished after seeking interesting comparisons between the characterisation of Ariel from Shakespeare’s The Tempest and the anthropomorphisation of AI models both in fiction and in reality. In Unit 3, I have expanded on this idea of anthropomorphism and found further correlations between the usage of emerging technologies and spiritual practices.
While reading through this course pathway’s Arena Research Platform, I came across a number of articles and sources that engage in the topic of ‘Technology and Magic’. One of these sources was the essay: ‘Ancient animistic beliefs live on in our intimacy with tech’ from Aeon magazine, written by Stephen T. Asma, Professor of Philosophy and Distinguished Scholar at Columbia College Chicago. Asma defined animism as having two key features: ‘one, belief that there are ‘agents’ or even persons in natural objects and artifacts (and even geographic places); and two, belief that nature has purposes (teleology) woven throughout it.’ (Asma, 2020) Ariel, from The Tempest, is a clear agent of nature, a spirit of air repurposed for Prospero’s revenge in the play, after being imprisoned in a tree. Asma indicates how: ‘we’re learning that humans can bond, form attachments and dedicate themselves to non-conscious objects… with shocking ease.’ (Asma, 2020) This view is later contrasted with Asma’s suggestion that tech-animism; bonding with technology, with spiritual respect for AI, might be an exercise in human sympathy: ‘Instead of dehumanising us, this tech-animism could actually be keeping us human.’ (Asma, 2020) The bond between AI models and human beings nevertheless highlights fluctuating power dynamics between the creator and the creation, the director and the performer.
‘Gods and Robots: Myths, Machines and Ancient Dreams of Technology’ written by Adrienne Mayor describes the concept of ‘black box’ technology as ‘machines whose interior workings are mysterious’ (Mayor, 2018, pp.3), comparing this to human intelligence: ‘It is often remarked that human intelligence itself is a kind of black box’ (Mayor, 2018, pp.3), and AI intelligence, in turn: ‘we are entering a new level of… black box technology: machine learning… will allow Artificial Intelligence entities to… select, and interpret massive sets of data to make decisions… with no human oversight’. (Mayor, 2018, pp.3) The social and political anxieties that surround the esoteric production of advanced technologies today has been compared in my artwork, and in my submission for the Research Festival, with the mass hysteria that had blossomed around the concept of witchcraft and the occult during Jacobean times.
Both subjects revolve around the fear of the unknown, around supposed sources of great spiritual and creative power. Both concepts offer potential threats to individuals and communities. Both are additionally historically delegated and dictated by patriarchal and colonial powers. Colonial forces of the ‘New World’ used the concept of witchcraft to criminalise and banish threats to their imposed systems of oppression, while patriarchal and colonial establishments today control and influence the AI datasets that render many artists powerless, stealing their works for generated content. My artwork therefore depicts an ongoing battle for the purposes and portrayals of witchcraft and AI, between the artistic individuals who seek to repurpose these practices, and the public dictators of these concepts in connection to governmental systems that stifle resistance. The animistic, demonic AI model of Ariel is caught in this conflict, though it remains an entity in my work that embodies corruption and anxiety.
Ariel, AI and The Witch’s Familiar
My practice surrounding Ariel as an AI model in Unit 2 transitioned in Unit 3 to encompass a narrative following the figure of the human creator and user. From this perspective, Ariel had become a presence symbolising uncertainty and danger, something that is acknowledged in referencing a witch’s familiar. Shakespeare’s Demonology: A Dictionary, written by Marion Gibson and Jo Ann Esra defines a familiar as: ‘a demon (or) spirit that attends on a witch or conjuror. The familiar spirit is often thought to be a personal servant to his or her human being, although familiars were also portrayed as shared…’ (Gibson and Esra, 2017, pp. 83 – 84) The dictionary continues to describe how the word ‘familiar’ implies a friendly companion, but this is contrasted with the following acknowledgement that in Jacobean times: ‘…demonologists feared that what was actually happening was different: instead of obediently submitting, they thought that familiar spirits would try to make a verbal or written pact with a witch… for his or her soul’ (Gibson and Esra, 2017, pp. 83 – 84). I was intrigued to discover this description as I found it could easily encompass some of my own personal anxieties when reflecting on my practice of working collaboratively with AI models. This was especially true in regard to my process when using Stable Diffusion AI generation software, that requires the user to download an AI models file and to grant permissions for the software to access your google account. The process additionally recommends the user only use the software if they know what they are doing. For me, this meant a relinquishment of control to the AI and to the black box technology which I only partially understand, even now, through my own research.
The Time Chat GPT Frightened Me By Describing My Own Art Practice To Me.
PROMPT: Could you help me to prepare a performance art piece by giving me commands that involve movement, sound, or both? It could be a choreographed dance routine, spoken word performance, or a musical improvisation. Tell me to embody a specific character or concept related to AI and ‘Ariel’ from ’The Tempest’.
ChatGPT Response:
Certainly! Here are some commands for a performance art piece inspired by AI and the character Ariel from “The Tempest”:
- Movement:
- Glide gracefully across the stage, embodying the fluidity of data flowing through a network.
- Perform quick, precise movements that mimic the computational speed and efficiency of AI.
- Flutter and twist like a digital sprite, representing the ethereal nature of Ariel.
- Sound:
- Create a soundscape using electronic and synthesized tones, symbolizing the inner workings of AI.
- Incorporate spoken word passages that explore the themes of artificial intelligence and freedom, drawing inspiration from Ariel’s desire for liberation in “The Tempest.”
- Experiment with vocal effects, such as echo or distortion, to evoke the digitized voice of AI.
- Character/Concept:
- Embody the dual nature of Ariel as both a powerful force and a servant, displaying moments of strength and vulnerability.
- Express the concept of AI’s boundless potential, yet the limitations imposed by human control.
- Convey the internal conflict of Ariel, torn between loyalty and the desire for autonomy.
- Choreography:
- Use synchronized movements with sharp angles and robotic precision to represent AI’s mechanical nature.
- Incorporate flowing, organic movements to portray the evolving and adaptive qualities of AI.
- Transition between contrasting movements, reflecting the contrasting elements within the character of Ariel and the nature of AI.
Remember, these commands are just suggestions to inspire your performance art piece. Feel free to adapt and modify them according to your vision and artistic interpretation. Good luck with your performance!
My Thoughts:
I can see aspects of my own description of my artistic practice in this response. I have highlighted in bold and italics the parts of the response that appear to resonate most with my artistic practice and writing. I presented this prompt in June 2023 when trying to consider what I would perform for my live performance in the Summer Exhibition. At this time, I had read that Chat GPT’s data is harvested from sites up to the year 2021, so it would be seemingly impossible for data from this website to have been harvested. I believe because it is free and it says so on the source, I have been using Chat GPT version 3.5. More perplexingly, the prompt I gave was in a new channel conversation/session with Chat GPT. While various online sources including Open AI’s developer forum claims Chat GPT has no memory of previous user conversations, this particular interaction felt especially eerie and appeared to defy this logic.
My wording in the prompt: ‘a choreographed dance routine, spoken word performance, or a musical improvisation… embody a specific character or concept’ was taken from a previous Chat GPT generation in which I asked Chat GPT to present ideas for types of artistic performance; these were the types in this exact wording that were listed. It reminds me of computer coding language – the idea that computers and AI have their own language, separate from a way that humans understand certain concepts and modes of conversation.
It seems that while AI is a reflection of the behaviour and information held by human beings, I also took on a form of AI reflection – repeating the wording of Chat GPT in my prompt. This appeared to somehow open a link between my thoughts and the response of the AI, leading to a Freudian uncanny feeling: a combination of the familiarity of my thoughts, my practice and my writing, but generated back to me through the unfamiliar presence of an AI.But is that even what had happened? Or is it all a mind trick that human beings play on themselves? When I read Chat GPT’s response, I was filled with an unsettling combination of excitement, confusion and discomfort.
Initial Reactions:
Bibliography:
Asma, S. T. (2020) Ancient animistic beliefs live on in our intimacy with tech. Aeon. [online]. Available from: https://aeon.co/ideas/ancient-animistic-beliefs-live-on-in-our-intimacy-with-tech [Accessed: 29th October 2023].
Bridle, J. (2023, revised edition. fp. 2018). New Dark Age : Technology and the End of the Future. London, Verso.
Butler, D. (2016). A world where everyone has a robot: why 2040 could blow your mind. Nature. Vol 530. February 2016. pp. 398 – 401. [online]. Available from: https://www.nature.com/articles/530398a
Centre for Artistic Inquiry and Reporting (2023) AI Open Letter – CAIR. [online]. Available from: https://artisticinquiry.org/AI-Open-Letter [Accessed: 29th October 2023].
Chadwick, J. (2020). Number of digital bits will overtake the number of atoms on Earth within 150 years with the world becoming ‘mostly computer simulated’ by 2170, study warns. Daily Mail. 12th August. [online] Available from: https://www.dailymail.co.uk/sciencetech/article-8615541/Number-digital-bits-overtake-number-atoms-Earth-2170.html [Accessed: 29th October 2023].
Gibson, M. and Esra, J. A. (2017). Familiar. Shakespeare’s Demonology: A Dictionary. [online]. Available from: https://www-dramaonlinelibrary-com.arts.idm.oclc.org/encyclopedia-chapter?docid=b-9781472500403&tocid=b-9781472500403-article-0000064&pdfid= [Accessed 29th October 2023].
Linebaugh, P and Rediker, M. (2000). The Many-Headed Hydra: The Hidden History of the Revolutionary Atlantic. London, Verso.
Mayor, A. (2018). Gods and Robots: Myths, Machines and Ancient Dreams of Technology. New Jersey, Princeton University Press.
OpenAI. (2023). ChatGPT (Version 3.5). [online]. Available from: https://openai.com/blog/chatgpt [Accessed 29th October 2023].
Shakespeare, W. (1999). The Tempest. Vaughan, V.M and Vaughan A. T. (ed.) London: Bloomsbury.