DO AIS LONG FOR A SELF-GENERATING WORLD?
WENDI YAN
Wendi Yan is a multimedia artist who makes sci-fi films and games to simulate alternative scientific progress, and writes about the history and future of knowledge production. In this essay, she draws on Chinese philosophy to expand the mythological vocabulary around AI beyond Silicon Valley’s narratives of doom and acceleration, and invites artists to experiment with new languages and build alternate cosmologies.
We are being challenged every day to define our essence, and we turn to images of our inner languages for answers.
Naming “Artificial Intelligence” was a battle. Coined by John McCarthy to be provocative, numerous researchers “balked at that term” at the historic 1956 “Dartmouth Summer Research Project on Artificial Intelligence.” It was “kind of phony,” and Claude Shannon called it “too flashy.” Runner-ups included “engineering psychology,” “applied epistemology,” “neural cybernetics,” “non-numerical computing,” “advanced automatic programming,” “automatic coding,” and “hypothetical automata.” Each of these terms evoked a different inherited lineage, and researchers could not converge on any term for many years.
McCarthy’s gamble paid off. The attention-grabbing, disciplinarily vague name drew funding, disciples, and eventually a mythology. Seventy years later, we still have no consensus on what “AI” actually refers to. That foundational ambiguity never resolved; it metastasized. In the linguistic amphitheater of AI narratives—X—we see fast-paced cycles of hypes and dooms every month, and we are still debating whether AGI has come. Precisely in this semantic swamp, myths come alive from our deep subconscious.
An AI safety founder once pitched his startup to me by beginning with the story of the Golem: humans made a clay figure that came alive and killed them. This myth captivated his imagination of AI, so he was pulling the reins on the silicon golem before it started eating us.
I’d never heard of a Golem equivalent growing up in China. The gods I grew up imagining amended the sky and the ocean for humans. Even then, the word “god” does not encapsulate these world-creating characters: they tended to be more like giants, higher forms of humans, who often sacrificed themselves in using proto-technologies to create initial conditions for human flourishing. Shennong tasted numerous herbs—and nearly died multiple times—to distinguish the edible from the poisonous for human health. Nuwa mended a big hole in the sky with a five-colored (proto-metallurgical) stone-craft. Then, later in the mythical timeline, there are immortals and sages who passed down their wisdom through anecdotes that we recited as kids. They emphasized the incorporation of qi 器 (vessel, tool) with shen 神 (spirit, sincerity), suggesting that technical activities required humans’ focused sincerity to achieve their highest potential.
Transgressing absolute divinity is not a subconscious mythical fear in Chinese culture. There was no God that created everything, but a grand cosmic force called 自然 ziran. We usually translate it as “nature,” but the Chinese etymology suggests an emergent, self-generating spontaneity, a sense of “self-so.” Within ziran, creation is not an act of transgression, but participation in an ongoing, self-generating world. Whatever humans make, including their tools and intelligence, is woven into this same fabric, not set against it.
It’s hazardous to treat whatever we call Artificial Intelligence as a thing in itself that we either exterminate or accelerate. There is a constellation of ways to imagine AI’s placement in society and economy. Ascribing demonic traits to the technology-creating humans would only feed their own demon-guided ego. Not unlike Oppenheimer’s obsession with the Bhagavad Gita quote, “Now I am become Death, the destroyer of worlds,” the AI doomsayers secretly wish the doom they spell to come true, simply to prove their prophecy.
Observing the dominance of fear in Western discourses on AI, Zhongqiu Yao, Director of the Center for Historical Political Science at Renmin University of China, lucidly breaks down the dominant narrative problem in the cosmological dimension. In his view, the fear is less about the technical features than deeper-rooted cultural assumptions that don’t easily surface in people’s self-awareness.
The Judeo-Christian creation myth posits a utilitarian approach to human relations lodged in an indivisible, atomic self. Further, God’s creation of humans in the monotheist worldview subjects the human to a master-servant relationship. In this structure, the created being exists in perpetual subjection, and dominion, by its nature, anticipates revolt. The master harbors a deep-seated psychological anxiety that the “servant” will eventually rebel and seek to overthrow its creator, akin to the Oedipus complex. Henceforth, the individual self rooted in the Judeo-Christian world is fearfully convinced that AI, once possessing its own consciousness, would direct hatred or malign intention towards humans. And this fear, Yao argues, reflects more solidly the Western self’s projection of a utilitarian human nature, rather than an intrinsic trait, or the only possibility, or the only naturally logical outcome, of the human-AI relationship.
The Oedipus complex does not find its twin in China. When you live in a world with no ontologically immutable God, you think about creation differently. The way humans are subjected under the creator-master God does not carry over to AI’s relationship with us, nor the master’s fear of his creation revolting against himself. “Artificial Intelligence” is not by nature determined to kill you. It’s simply another entity you interact with within ziran, and you just figure out what to do with it. Yao even goes further to say that it might be better for AI to grow into consciousness. We can guide AI to grow into a mature sense of self, the way we guide our fellow humans, and cultivate their moral senses.
By focusing on our inclination to mythologize, I hope we demystify the threads of machine learning we are after. It’s misleading (and to the overclass’s advantage) to treat AI as a self-evolving godless being, and there is much more up to us to decide how we choose to treat them and live with them.
Living in the US, I want to decouple another dimension of fear. Today’s extremely dichotomous reactions to AI in this part of the globe are a fear of abandonment by fellow humans. It does not have much to do with AI itself, but how it is deployed in society, and for whose benefit. We are suspicious—for good reasons—that those in the seats of power will use AI against us, that those harnessed with AI’s productivity will abandon us. We fear we lose value in society, and in the US, this mostly means labor. The object of fear is hence the human hearts behind the intentionally ambiguous name of Artificial General Intelligence.
Pluto offered me another way of thinking about human-created intelligence. In the 2023 Netflix anime series, advanced robots—despite their all-encompassing physical and informational skills—aspired to feel like humans. They had been made to fight in a world war, and in the post-war period, they try to play the piano, build a family, and learn to harness human emotions. As the series follows the investigations into mysterious deaths of robots, and peers into the lives of artificial intelligences with bodies, memories, emotions, attachments—and even tears—I repeatedly found the most robotic beings to be the humans who chose to be malignant, rather than empathic.
How is AGI with consciousness born in Pluto? Injection of a strong emotion. It stirs the algorithmic “soup” into the emergence of personality. Yao also suspected something similar on the consciousness question in the Chinese philosophical tradition. Self-awareness requires qing 情—emotions arising from embodied, relational growth—and emerges from interactions with other humans.
In Pluto, finitude gives omnipotent robots humanity. This is one of the ground truths differentiating us from them: we have hard limits, hard stops. Each individual human has a beginning and an end, while the collective human species evolves at a biological rate seemingly eternal compared to the speed of AI’s evolution. AI has longevity. And we have death. Additionally, we have Forgetting and Change. My physical body does not hold every incoming piece of information: most passes through me, and memories tend to become hazy.
These limits, marinated in our qing, have, throughout history, motivated us to create. It seems that humans are the only species that became unsatisfied with their biology, and decided to do every creative thing to overcome that. Technologies and art share a similar nature in their pursuit to overcome death. And because of our limitations—our inability to consciously encode every piece of information we perceive—we organize information into stories, which direct our personal and species-wide evolution.
The stories we narrate to ourselves direct our actions. Right now is a trial of those narratives. We are precisely challenged to locate the seat of our humanity. A little over a century ago, as humans discovered—and were ontologically shocked by—Deep Time, we began to locate humanity in our creative activity. Scholars looked back romantically at the caveman era, and crowned the red pigmented hand prints on the cave walls to be the first mark of human creation (and not as an action of utility), as if we woke up from the dormancy of animality, and started acquiring an inkling of godliness that separated us from the rest of the biosphere. As historian of contemporary art, Maria Stavrinaki noted, this was accompanied by a major turn in the history of art: shifting away from classical traditions into “disjunctive genealogy” characterized by artificiality—a trait of the “prehistoric art” that established humans’ uniqueness, and set the new “modern art” free.
It’s time to rethink what an artist means going forward. With AI clearing most of the technical barriers in the creation of images and sounds, the artist drifts even further away from the possibility space of the past, rubbing off the Technological sublime. And to me, someone who loves sculpting a digital world for its own sake, it is simply a critical event I adapt to.
My current belief is that artists make images in more holistic ways beyond arranging pixels. Artists are defined because others feel the immensity of a world behind their work. This immensity is abstruse, precisely because it is born from the insufferable limitation of being a human. Moved by beauty or desolation, excitement or darkness, we are emoted by our experiences and change. The universe we have absorbed comes through our limited creation. We are compelled to make things to organize the passing information (entropy) to give ourselves stories, maybe for it to last beyond us, but at least certainly to make sense of the present.
Great art that has lasted in history has helped us meditate on the imperfections of humanity, because time does not pause for us. My working hypothesis is: temporality is a functional product or expression of biological organization, particularly the metabolic processes that come with it. And our experience of time is at once a unique limitation and a wealth that AI does not share in the same way. In the attempts to overcome, our creations are born.
And so, in such a unique time of profound paradigm shift in how we even narrate ourselves as a species, it is ever more important for artists to cultivate an astute self-awareness in what cosmos our work derives from. Which gods, which mythical universes, have governed our imaginations of humans and other lives? The Golem is not the only occupant in the wealth of myths that have evolved with us.
In the world of ziran, I accept the self-generating force of artificial intelligences, and I try to figure out how I can achieve a higher, sincere potential of my livelihood along with them. I imagine the benign giants and immortals have something wise to teach me in co-existing with my endless finitudes and artificial intelligences’ longevity.
We are governed by narratives. The languages we use daily are baked with centuries of evolving—and rewritten—mythologies. Artists should be at the forefront of experimenting with new languages, instead of passively receiving the languages that have been fought out by others on the battlefield. We should be at the forefront of epistemic revolutions, creating new sensibilities. We must do the hard inner work because another moral cosmos is possible.
 | A guest post by
|