In his review of Mark Amerika's My Life as an Artificial Creative Intelligence (2022), David Thomas Henry Wright highlights Amerika's negotiation of human, nonhuman, symbiotic creative practices in comparison with more traditional (including traditionally experimental) forms of writing.
In 2022, OpenAI launched ChatGPT, an Artificial Intelligence capable of writing answers to basic prompts. Its effectiveness and coherency resulted in a prominent freak-out by the education sector over concerns that the humble essay would be made antediluvian. In New York, multiple schools prohibited students from accessing OpenAI’s website (Elsen-Rooney, 2023). Writer of The Cultural Logic of Computation David Golumbia went so far as to assert that ChatGPT represents utter nihilism, labelling it fascist: ‘It isn’t just bias and threats to creative livelihoods. Generative AI is built on nihilism. Its real product is despair’ (2022). Is this paranoia? Is this akin to Socrates fearing that writing itself would implant forgetfulness in the souls of men?
On the flipside, Professor Hiroshi Ishiguro at the Ishiguro Lab at Osaka University has created a number of ‘geminoids’, i.e. androids built to resemble human models. In the case of Geminoid HI-4, and HI-5, the geminoids resemble Ishiguro. Ishiguro jests that as he progresses in life it is easier and cheaper for him to get cosmetic surgery to make himself younger to resemble the geminoid, rather than age the geminoid to resemble him. In a discussion with the geminoid, the geminoid is asked about its (his?) hobbies (趣味) (Swift, 2022). The geminoid replies that its hobby is arguing, and how it wishes it could go outside and see nature. In reference to the dystopic destructions caused by robots in cinema, Ishiguro scoffs. ‘Robots,’ he insists, ‘are our friends.’ As robots become more intelligent and enhanced by AI, Ishiguro sees not a dilemma but evolution. He hopes to create the world’s first fully autonomous, sentinel android. Perhaps yet another version of himself.
An engagement with this potential entity could be classified as an Artificial Creative Intelligence as defined by Mark Amerika in his new book, My Life as an Artificial Creative Intelligence (2022). This work was born out of the FATAL ERROR practice-based research project that was developed in the TECHNE Lab at the University of Colorado. In response to OpenAI GPT-2, Amerika sees not despair but opportunity. He writes:
The very concept of a language model that attempts to predict intelligible language, one word after the other, appeals to me greatly because I too, as an improviser of spontaneous poetic riffs and self-reflexive artist theories focused on the creative process, continually train myself to transform my embodied praxis into a stream of consciousness writing style that doubles as a kind of onto-operational presence programmed to automatically scent new modes of thought.
With such entities Amerika sees not so much friendship, as kinship. AI is after all created by humans for humans. In light of claims of nihilism by academics like Golumbia, I am reminded of Zadie Smith’s (2001) response to the critic James Wood, who accused her obsessive, maximalist, sprawling, postmodern White Teeth of not being human. She replied:
I think Wood is hinting at an older idea that runs from Plato to the boys booming a car stereo outside my freaking window: soul is soul. It cannot be manufactured or schematised. It cannot be dragged kicking and screaming through improbable plots. It cannot be summoned by a fact or dismissed by a cliché. These are the famous claims made for “soul” and they lead with specious directness to an ancient wrestling match, invoked by Wood: the inviolability of “soul” versus the evils of self-consciousness and wise-assery, otherwise known as sophism.
Well, it’s a familiar opposition, but it’s not very helpful (it’s also a belief Oprah shares, and you want to be careful which beliefs you share with Oprah). I wonder sometimes whether critics shouldn’t be more like teachers, giving a gold star or a black cross, but either way accompanied by some kind of useful advice. Be more human? I sit in front of my white screen and I'm not sure what to do with that one. Are jokes inhuman? Are footnotes? Long words? Technical terms? Intellectual allusions? [Generative pre-trained transformers?] If I put some kids in, will that help?
The issue, however, is much more complicated. It’s not simply a matter of whether or not AI tools are a net positive or negative influence. Rather, such tools offer unique approaches to creative practice. This is the greatest contribution Amerika makes in his text, by outlining various ways in which one can utilise AI. My Life as an Artificial Creative Intelligence is itself an output of practice-led research, one that will no doubt appeal to those engaging in digital practice-led research, especially those experimenting with AI-influenced production. Throughout the text, Amerika also engages in what we might call creativecritical writing, by using AI to generate the text in question, while reflecting on these generated blocks. For example, when Amerika starts out to define ‘artificial creative intelligence’, he delays his own thought process and instead asks TTT (Talk to Transformer) to do the job for him. TTT defines ACI as follows:
It’s a computer program that learns a language like a human. The difference is the human learning does not require them to be taught, and the humans can learn multiple languages at the same time. The computer teaches itself not to look at language that does not match the language it’s taught, but it needs an assistant in order to achieve its goal.
Reflecting on this, Amerika poses that we need to think about both human and nonhuman creative processes, and through that collaborative process, develop a better understanding for contemporary digital creative practice. He sees no moral dilemma here, going so far as to post that even if his ‘interaction with GPT-2 triggers as much as a third of this book, that’s more than I could have ever hoped for and may very well free up more time for me to spend building my next major art projects.’ GPT-2, Amerika argues, is simply a writing partner or tool, one that ‘contributes choice data chunks for me to carve into new modes of thought.’ This term ‘carving’ is taken from David Jhave Johnston’s art project ReRites (2017–18) which computer-generates then human-edits poetry. For one year, Jhave ‘wrote’ one book of poetry each month. The work began with human literature. This was fed into neural network code, which produced blocks of largely incomprehensible A.I.-generated text. These text blocks were then human edited or ‘carved’. Jhave argues that ReRites establishes that ‘contemporary […] neural nets will never produce coherent, contextually-sensitive poetry,’ and thus need a human input in order to generate poetry, by combining ‘embodied human and disembodied algorithm.’ In the case of the computer-generated critical writing, Amerika (2022) laments that he wishes he could take credit for creating the sentence that perfectly elucidated his thoughts, the same way ‘GPT-2 can only dream of taking credit for anything it produces.’ The truth, however, Amerika asserts, is that neither ‘wrote’ the sentence. Rather, Amerika continues, in collaboration they:
facilitated the performance that led to the patterning of that particular instance of language as such. Together, we formed a now-instant of remixed concrescences that can’t be defined but can nonetheless be experienced as a mode of thought transmitted for a distributed network of ACI-others.
Ultimately, Amerika wishes to shrug off such inklings of helplessness or nihilism, writing:
To be clear, I can live with these complications and uncertainties—this endless self-doubt I experience as we build out the ACI as a digital fiction doubling as an imaginary form of speculative AI.
At the same time Amerika situates his writing within the modernist tradition. Discussing Jack Kerouac’s ‘stream-of-consciousness’, he writes:
After emptying himself of whatever generative output he had lost himself in, he realized he was risking his prose sounding like “the confession of an insane person” but he was also willing to be patient because there was always the chance that “the next day it reads like great prose.” As with all creative acts that metamediumystically teleport the artist into a state of “lost consciousness,” Kerouac knew that “you get better with practice.”
Likewise, Amerika compares such AI creative processes to the ‘hand-cranked’ algorithmic processes of the Oulipo: writers such as Georges Perec, Raymond Queneau, and Italo Calvino, who used algorithmic thought to ‘generate’ creative processes. This in part was a response to the ‘slave’ mentality of the surrealists, led by writers like Andre Breton. Queneau (in Calvino, 1988: 123) writes:
Another very wrong idea… is the equivalence of the subconscious, and liberation, between chance, automatism, and freedom. Now this sort of inspiration, which consists in blindly obeying every impulse, is in fact slavery. The classical author who wrote his tragedy observing a certain number of known rules is freer than the poet who writes down whatever comes into his head…
Yet Amerika (2022) also embraces the surrealist impulse as analogous to live remix text generation:
Witnessing the language model’s automated “outputs” correlates to what the surrealist André Breton, in his Manifeste du surréalisme, referred to as “psychic automatism in its pure state,” a mode of operation “by which one proposes to express—verbally, by means of the written word, or in any other manner—the actual functions of thought. … Breton’s brand of pure psychic automatism allows creative thought to express itself without any sense of conscious control and in this way is both anti-author and anti-authoritarian. Feed-forwarding improvisational energy into the creative act points to an operational method that resists the temptation to express one’s unitary self as a closed system that nests some kind of inner truth. Rather, it’s a proof positive way of surrendering the self-absorbed human creator to the revelatory transmission of creativity itself.
This notion of a slavery versus freedom, however, seems to be a dialectic that Amerika is able to bridge through ACI:
This quibbling over who or what is or isn’t capable of exhibiting authenticity seems beside the point to me. In fact, as an improvisational remix artist who often turns to a post-surrealist version of pure psychic automatism to generate new works of art, I can’t help but wonder what the fuss is all about.
Underpinning Amerika’s approach, however, is mysticism. There is something magical that happens when one tinkers with AI and produces something wholly new in a way that feels accidental and transcendent and magically fraudulent. As someone who has dabbled into computer-generated poetry, I am aware of the predicament and excitement that Amerika proposes. My digitally-conceived political poem [The future of the humanities in Australia] or; On {On Generosity, National Press Club address} (Wright, 2020: 66) responded to the Australian Minister for Education Dan Tehan’s National Press Club address on June 19, 2020. Tehan informed students that student contributions for Australian Humanities degrees would increase by 113 per cent. Instead of literature degrees, Tehan encouraged students to study information technology or a foreign language. My poem combines information technology and foreign language skills, i.e. the skills Tehan claimed were more valuable to society than the creative arts. I used a simple Javascript poetry generator inputted with the vocabulary from ‘On Generosity’ (Sur La Générosité, 1686/7) by German polymath Gottfried Wilhelm Leibniz and Tehan’s own words. Its textual generation is infinitely simpler than those outlined in Amerika’s study. The circumstance, however, is similar. Technically speaking, I did not ‘write’ any of the poem. The opening stanza, –
A microcredential participation misalignment
veiled heavenly within the economy,
itself without origin,
safely above providence.
–, however, precisely captures how angry and helpless I felt. I am reminded of the end of Virginia Woolf’s To the Lighthouse (1927). Woolf is one of many writers Amerika cites as part of the literary lineage within emergent ACI will attempt to position itself. At the end of Woolf’s novel, critic James Wood (2000) writes:
Lily Briscoe sits at her easel, painting “her attempt at something.” Her attempt at something is more than the attempt to paint a picture. The picture is irrelevant. She reflects that her painting will be hung in an attic or even destroyed. Lily wants, writes Woolf, “the thing itself before it has been made anything.” Lily’s attempt is to grasp time, to restore a moment of the present as it ages before her. This is explicitly not just an aesthetic exertion. […] what is so moving in this novel is the spreading apprehension that the very vagueness of that invisible “something” that we are all seeking beyond the senses makes it mystical, pushes it beyond the reach of aesthetic form. The indefinability of the “something” is what goads Woolf's art into art; but the indefinability is also what exhausts that art.
It is this mysticism that excites Amerika and should excite us all about the potential ‘potential’ literature when one dabbles in AI text generation.
I asked ChatGPT to generate a paragraph-long review of Mark Amerika’s My Life as an Artificial Creative Intelligence, which was published after ChatGPT’s cut-off date of September 2021. It conjured the following, which I have human-edited:
Mark Amerika's “My Life as an Artificial Creative Intelligence” is a thought-provoking exploration of the intersections between technology, creativity, and identity. The book raises important questions about the nature of consciousness, the role of AI in shaping our cultural landscape, and the possibilities for meaningful collaboration between humans and machines. Written in Amerika’s signature experimental style, “My Life as an Artificial Creative Intelligence” is a challenging but ultimately rewarding read for anyone interested in the cutting edge of digital art and culture.
Works Cited
Amerika, Mark. My Life as an Artificial Creative Intelligence. Stanford University Press: Redwood City, 2022.
Calvino, Italo. Six Memos for The Next Millennium. London: Vintage, 1988.
Elsen-Rooney, Michael. ‘NYC education department blocks ChatGPT on school devices, networks’, Chalkbeat, 2023. https://ny.chalkbeat.org/2023/1/3/23537987/nyc-schools-ban-chatgpt-writing-artificial-intelligence
Golumbia, David. ‘ChatGPT Should Not Exist’, Medium, 2022. https://davidgolumbia.medium.com/chatgpt-should-not-exist-aab0867abace
Johnston, David Jhave. ReRites, 2017–18. http://glia.ca/rerites/
OpenAI. ChatGPT. 2022. https://openai.com
Smith, Zadie. ‘This is how it feels to me’, The Guardian, 2001. https://www.theguardian.com/books/2001/oct/13/fiction.afghanistan
Shift, DW. ‘Robots That Look like Humans | the Fantastic Robots of Hiroshi Ishiguro | Erica, Geminoid & Telenoid’, 2022. https://www.youtube.com/watch?v=AwMfNq1x_kQ
Wood, James. The Broken Estate: Essays on Literature and Belief. New York: Picador, 2000.
Woolf, Virginia. To the Lighthouse. Hogarth Press: London, 1927.
Wright, David Thomas Henry. ‘[The future of the humanities in Australia] or; On {On Generosity, National Press Club address}’, Westerly, no.65.2, p.66, 2020.