AI and Apophatic Re-Humanization
What will be left when the whole world is automated?
Recent advances in AI, and especially AI image, text, and sound generation, have provoked a renewal of the perennial anxiety that machinery is eclipsing or obsolescing humanity. The fact that this latest wave of advances has targeted the arts instead of engineering or manufacturing has more specifically provoked uncomfortable questions about the nature and purpose of art, and especially the way art has for decades been one of the last territories in which a transcendent conception of human nature is not marginalized. Genetics, chemistry, biomechanics, and even psychology and sociology have all been tightening a noose around the question of what a human being is. Homo Sapiens, they say, is a pattern of nucleic acids and minerals which acts according to patterns of electric and chemical stimuli, patterns which are informed and deformed and dysregulated by nutrition, habitat, epigenetic inheritance, and semi-autonomous models of socialization that overlay these things. Will keeps retreating from us, to the point that hardly anyone bothers to seek it anymore.
The long, grinding demoralization campaign of midcentury scientism has utterly desacralized the human body and brain. The response to this, for many, has been to outsource one’s consciousness to a phone, iPad, or PlayStation. Post-Y2K generations do not grow up with an embodied consciousness, they grow up with an electronically distributed consciousness that is periodically intruded upon by a decaying, PUFA-ridden body. Interestingly it is these generations in which anti-AI sentiment seems strongest. My 55 year old mother loves AI art; for her it is a vehicle for creating interesting and appealing things. But for a very online millennial, the habitat of most of one’s conscious experience is the internet, and the only thing which lessens the crippling loneliness of that habitat is the reassurance that the content consumed is coming from other humans. AI threatens to take that from them, and anxiety about it in a way is a crisis of identity. But this Band-Aid has already been scratched at; perceptive people should have already seen that the selective pressures of the algorithms and the outsized influence of state and corporate propaganda has made the internet overwhelmingly fake for years now. There are no humans online, there are only homunculi, tulpas, and golems.
AI cannot take your humanity away from you. But, just like every technological revolution before it, it will obliterate structures which you have invested your sense of humanity into. This might feel like a lot like losing your humanity, and a lot of people might ruin themselves over it. But ultimately there is nothing new under the sun; One’s humanity is found by a stripping-away, by a distillation, not by scrambling for new shells like a hermit crab.
Before Jane Goodall’s research into chimpanzees, it was believed that humankind was categorically different from animal-kind and that the usage of tools was the arbiter of this distinction. Then chimpanzees were seen using tools in the wild. After that the arbiter of humanity was the use of language, and then numerous great apes and even some birds were proven to be able to use language and some animals have even been shown to use para-linguistic codes in their communication in the wild. The old need to scientifically prove our compulsive sense of superiority over nature shows that our existential anxiety once cut in the other direction.
But since then, the scientific consensus is basically to admit defeat. There is nothing un-natural or super-natural in man, and now we believe we have mimicked nature in creating the proper arrangement of dead matter and electric currents for “intelligence” to exist. So long as a materialist episteme is maintained there can be no solid argument made for why humanity should continue to live in a body or maintain the earth’s biosphere. And for this reason, the survival of humanity depends on the destruction of this scientific paradigm and especially of its approach to organisms as being mechanistic. Ironically, “trans-humanism” can only possibly have any validity within a metaphysical frame that could permit or facilitate the migration of a continual awareness into a non-biological body, perhaps somewhat like how the Egyptians conceived of animating an eidolon.
As things stand now, AI anxiety really comes down to the fact that we are rapidly realizing that discursive thought is mundane, a property of physical nature, and that we are not even particularly good at it. This is ultimately stressful to us for the sole reason that we have forgotten the fact that true awareness is cosmic, it is the spacious power that suspends all things in immediate presence to their causes, up to and including God.
But honestly, in my opinion, this whole kerfuffle is a little premature. “AI” currently really just means “language model,” i.e. a supercharged parrot connected to google and a dictionary. Even the image generators at this stage work by mixing and matching patterns of visual tokens. No matter how minutely perfected these methods will become, there can be no Will or autonomy in them and, most importantly, no endogenous valuation. GPT-4 does not have an awareness of words’ meanings outside of the statistical patterns in how humans use them. It cannot assign a “conscious” value to words, i.e. for GPT words are whole items in themselves and not tokens for shaping patterns of awareness the way they are for humans. If a language model ever makes a claim to be conscious or to have general intelligence, we can safely dismiss it on the basis that it is only mimicking what human speech patterns are statistically likely to say.
The radical flip side of this is that perhaps there is no hard line dividing biological bodies from artificial ones in the sense that the divine awareness could be present to all things regardless of what organs they possess for parsing it. Perhaps GPT-3, my childhood Roboraptor toy, and Earth’s tectonic plates all have a share of buddha-nature, or at least a share of the world-soul.
Regardless, there can be no more fleeing from metaphysics. The science that once cleaved man from nature is now beginning to cleave man from his own self-image. AI art is distressing to people because our 21st century self-image is one of a shriveling corpse who strives after immortality only through leaving behind works that surpass it. We view AI as threatening to outcompete our work and thus deprive us of this immortality. But 21st century man and his self image was already stilted, and to be cleft from it may be a blessing.
For the last few decades, the fact that art remains the last bastion of a supra-physical humanity has also made it a safe sandbox to retreat to while everyday life becomes more and more dehumanizing in ways that are impossible to ignore yet extremely difficult for normal people to recognize and articulate. Why do we feel it is unacceptable to lose the human touch in media we consume, but eagerly accept the loss of human touch for the physical products and even the food we consume, having so much of it delivered to us without seeing its maker, a cashier, or the delivery person who drops it off before you know they’re there? Why is convenience worth the price of disconnection with retail but not art?
I believe that people recoil from automated art because there is a vestigial instinct that human arts are materialized images of our souls, that an artist is a midwife for something transcendent, and that the beauty of a work is directly related to this process of creation which invests it with the presence of the archetypal Beauty. Again, as with my prior mention of the fake internet, art has already been extensively profaned and commercialized to the point that this panic almost verges on hypocrisy, but this discussion is still important to have for the sake of what it can reveal.
If automation divorces visual and audio art from the “psychourgic” activity of a human artist, then I believe the result will be much like what happened at the advent of photography and will have similar ramifications for how society thinks about imagination and about the contents of imagistic consciousness.
With the invention of photography, there was widespread panic that painting as an art form would die. Instead, painting simply surrendered the ground of realistic representation and began the shift towards abstraction, starting with impressionism and related styles. The objects of artistic depiction ceded their primacy and sovereignty to new methods of depiction and new contexts for their depiction that, when done best, could blur the subject-object distinction and evoke a direct experience of abstract universals; the west essentially re-invented pagan iconography without commensurately resurrecting pagan theology and psyche-ology to ground it.
When photography caused art to retreat into the conceptual and take advantage of its ability to depict abstracts, in a very real sense it pushed art to become more human, more related and attentive to the inner experience of humanity instead of a meticulously accurate representation of physical reality. If AI causes humans to retreat from the direct manual creation of sound and image, could an analogous shift be repeated? I believe that at that point what is left will not be a lack of art, but a new understanding of art: Without the trappings of method that tethered abstract art to the physical, the instant and almost seamless transition of an image from ideal to manifest will mean that the last limiting factor on art, and therefore the last substrate and substance for Mastery to emerge in, will be the imagistic consciousness itself. The master artist of the future may be somewhat like a shamanistic figure, one who, within a society of passive consumers, is uniquely able to be present to the world soul and to the universal, archetypal beauty and then convey that presence into linguistic or imagistic thought, at which point AI takes the reins for its physical manifestation.
Given humanity’s track record and the revealed preferences of average people in today’s age of relative abundance, I think it is a safe bet that for most of society AI will not actually change what kind of content is consumed. Most art will continue to be simple entertainment and escapism, purely appetitive. Whether this escapism is produced by the market, which is only an egregoric algorithm that uses human processors at the top instead of robotic ones, or produced by an individual person renting GPT-50’s processing power to make a movie, is almost irrelevant. Commercialization and marketization have practically already automated entertainment.
But for snobs and fanatics, the change in methods that AI could bring may be enough for a new conception of high art and maybe even a sort of apophatic renaissance: As art is forced to retreat farther and farther from representation, perhaps it is pushed closer and closer to sources and to universals. Photography changed art’s relationship to beauty from imitating what was already beautiful in the mundane world to attempting to portray what is beautiful in the inner experience; Now, AI stands to deepen that relationship by removing method as an avenue of mastery so that only the conceptual is left. If art can’t be judged by its mastery of imitation or its mastery of execution, maybe it can be judged by the strength of the ideas it depicts, and in order to judge it on these terms viewers will have to become familiar with how an idea can be beautiful in itself.
To become familiar with the beauty of an idea or with the ideal of beauty, one must become familiar with what it is in the human being that is capable of perceiving beauty. In other words, one must philosophize. The wonderful thing is that to find what it is in yourself that perceives and understands and is present to beauty is also necessarily to find that same thing in all people and in the cosmos itself. If AI generation can democratize the creation of art, there is at least a small hope that it will similarly democratize a new understanding of the artistic process and the purpose of art, an understanding which contextualizes art in terms of the total, microcosmic human being, including his metaphysical bits.
I have until now partly brushed over an important aspect of art, which is that it cannot be divorced from struggle. The latin Ars from which we derive “art” literally means “skilled work.” The striving after mastery is part of the beauty of any creation. And automation, more than anything, lessens struggle - at least outwardly. When there is no difficulty left in the physical conjuration of an image or a song, there is no mastery left to be gained and thus no matter how much beauty is in the work, that physical token is not necessarily art in the human sense but almost more akin to a natural landscape. Composition, painting, sculpting, building, and any other manual work can and will be automated. But it is absolutely impossible to automate philosophy and theurgy. Focused, controlled, practiced mental presence and attention are inherent to them. These are arts which concern the direct experience of the non-physical, the non-sensible. For these reasons they may become the last truly human arts, and truthfully they have always been the most human arts. To re-humanize ourselves, I believe, must mean to pursue them.
(Post-script - image generation is certainly far behind text, and I think we’re a ways off from any of this really coming to pass. To be topical, I wanted to use an ai image as the cover photo for the post but couldn’t get anything satisfactory out of three different generators. Maybe if midjourney was still free I could have wrangled something. Still, automation is a relevant topic in almost every part of our lives and the threats posed by AI do not differ that sharply from threats posed by the flesh-based algorithms of government, market, and pop “discourse.”)