BUT, IT’S BEAUTIFUL! Why Artificial Intelligence Can’t Make Art
These days, It seems like every day I stumble upon an article discussing an interesting new development in the world of artificial intelligence (AI). Each day holds a new advance, discovery or existential worry in regard to AI’s exponentially burgeoning relevance. Indeed, much of the AI-related media I have consumed contains the explicit notion that people are finding it difficult to keep up with the technology’s growth - including the developers. While there is a nearly infinite sea of interesting topics and ideas surrounding current AI developments, as an artist I cannot help but be drawn to the “art” generated by AI and the implications of its artificial creativity. Some of the images I have seen are fascinating to be sure, and the complicated processes that AI uses to create these images are just as interesting (although admittedly, often beyond my technical understanding). Luckily, I do not need a working understanding of the algorithmic processes that underlie AI’s images to appreciate them. But are these tech-generated images truly Art with a capital A? Or are they simply interesting visuals? This has been the subject of much debate, and arriving at a clear and rational conclusion may be difficult if we don’t all share the same, high-level definition of what art is. For ages people have battled the conceptual hydra of attempting to define art, and usually, the disagreements are matters of little consequence. Eternally lacking objective and definitive boundaries, the art world continues to press on creating. Occasionally, when an artist pushes the “limits” of creation and approaches that blurry edge of the imaginary boundary, the conversation sparks up again. Why should it have a definitive boundary? After all, is it not the very absence of concreteness that makes the world of art and creative expression what it is? Probably so. In the case of AI, however, lacking a conceptual definition for art may be putting a large piece of our humanity at risk, with far-reaching consequences.
• • •
Before determining whether AI “art” is art at all, we need a definition of art that is properly shored up. Simply put, I believe art can be defined in its most basic term as the intentional and tangible expression of experienced reality. Poems, paintings, songs and all other forms of creative expression communicate something about the internal world of experiential consciousness. For example, a beautifully crafted chair indeed says something about the creature that built it (fish do not make chairs), but it does not communicate anything about the crafter’s experienced reality. A landscape painting, however, says everything about the internal world of its sentient creator. Functionally, art is largely a form of expression that fills in the expansive chasms of the human experience for which language so often falls short. Human language, though immensely sophisticated, can communicate only a narrow spectrum of the human experience. This can be likened to the wide range of the electromagnetic spectrum from radio waves to gamma rays, from which our eyes can only detect a narrow band of visible light. Language reliably falls short of how we think, feel, and experience much of the time. Though I am speculating, this may explain at least in part why artistic expression appears after the advent of rudimentary language in the human timeline. Art exists for its own sake. It is a communication device created by conscious creatures, for conscious creatures with the intention of communicating some facet of what it is like to experience. This is precisely why art resonates with us so deeply: we connect and identify with the experiential, intellectual and emotional state of the conscious creature who created it. We understand what it is like to experience the world, and we feel that in the language of creative expression. We also value the novelty and originality of art. There is without doubt a drastic shift in how we understand, conceptualize, and emotionally resonate with a piece of art once we find out it is a convincing copy rather than the original. While we might agree that, for example, the copy demonstrates an advanced and impressive technical dexterity on the part of the plagiarist, the art was not born out of an authentic conscious experience. The only thing that could have birthed a specific original piece of art was the precise experiential and emotional state that the artist was in when he or she created it. That is what we truly long for in art. It is the connection with others experiencing the world… the universe… reality… just like we are. It is a way that by understanding others, we further may understand ourselves. Without that connection, an oil painting is simply an image, not a piece of art.
• • •
Despite what a very small minority of journalists and tech experts might speculate, none of the AI that has been created is conscious. While AI technology might be highly sophisticated in replicating a conscious agent, we can be quite certain that the lights are not on… yet.
Consciousness is indeed a giant mystery, perhaps the biggest mystery of all. Without knowing precisely what it is or exactly how it arises, hastily ascribing the miracle of sentient awareness to our latest technological invention is, at the very least, dangerous. An AI technology may be able to beat a human at a game of chess, but it is not aware that there is anyone “out there” that it is playing. It is not aware of its opponent or the universe in which itself and the other human player reside. AI is accelerating at an exponential rate but at the present time still falls short of human intelligence. And there is not a single grounded reason to believe it is conscious.
• • •
All life on Earth likely originated from a single, self-replicating molecule. Each new living form, generation after generation, shares some level of kinship with all other living creatures. Some are more closely related than others depending on how recently two organisms shared a common ancestor. Perhaps shockingly, humans share 99.8% of their DNA with chimpanzees. This means that a difference of a mere .2% within a vector of evolutionary development can foster the differences in behavior and intelligence between us and chimps. For some animals, a greater level of intelligence is an attribute of fitness that is selected for through evolution by Darwinian natural selection. When genes are replicated to construct a new individual, they occasionally make mistakes. These mistakes come in the form of genetic mutations which will occasionally have a beneficial phenotypic effect on the individual. For example, a genetic mutation that allows for an individual rabbit to have an abnormal increase in leg muscle will be able to outrun predators better than others. Evading predators more successfully and more often than other rabbits and prey animals means that it will live longer and have more opportunities to pass on its genes into a new generation. It is the dream of each individual gene in every body to exist in the world for as long as it possibly can. It is not difficult to imagine a lifeform, living on some distant and remote planet whose evolutionary environment allowed for the development of a significantly higher intelligence than our own – one that makes the difference between us and our closest cousins seem negligible. If the lifeforms were biologically organic, their level of intelligence must have been originally selected in their home environment as a matter of fitness through a process algorithmically analogous to natural selection and possibly sexual selection. What kind of environment might that look like? With so many variables, all of which are unknown, it is impossible to know. Suppose these lifeforms decided to visit our corner of the universe and stopped by to say hello, and that they were somewhere within our own range of intelligence so that an introduction was actually possible (not like a human trying to interact with an earthworm). What would we show them? In what ways would we display to them our “human-ness”? Would we show them our equations? Our telescopes? Our engineering skills? These creatures have the technology to successfully visit us from a dark and mysterious corner of the universe. Certainly, our technological and mathematical prowess will not impress them. In order to truly convey to the travelers our identity as a species and to display to them who and what we are, we would show them the best of the creative human spirit: we would show them our art. We would put on full display our highest creative achievements in music, painting, poetry, dance and sport…our humanity. We would be conscious creatures communicating to other conscious creatures.
• • •
Consciousness must necessarily play a role in both the creation and the consumption of art. Otherwise, there is nothing to communicate. Technological machines running on algorithmic processes do not think, feel or experience. Therefore, they have nothing to communicate about what it is like to be. When making a piece of art, an artist makes aesthetic choices that are, in-part, based on individual psychology and structured value hierarchies. In humans (as with all living creatures), much of what determines the particularities of these value structures stem from evolutionarily universal attributes that were naturally selected for in our remote ancestral past. Machines do not emotionally and psychologically value anything. Value in this context must necessarily involve experiential consciousness, otherwise, we are talking about a wholly different kind of “value.” I am afraid we lose a great deal of our humanity by calling an algorithmically-generated image “art”. I am certainly not making the claim that AI cannot generate interesting images. Even more importantly, I am not claiming that an AI-generated image or piece of music cannot emotionally move the human mind. I am certain that it could. However, we must remember that true artistic expression is about communication of our inner world – the human condition. We value art because we are social creatures who value each other and the experiential worlds which we all individually occupy.
• • •
The new wave of AI development has brought many important philosophical questions to the forefront of our society and culture. It forces us to open up a dialogue and reassess what it means to be human in a fundamental way, and that is a great thing. The issue of AI and art is just the beginning of a long range of crucial conversations that will need to take place if we are to safely push forward with these new technologies. In just 30 years we’ve seen society-changing innovations including the internet, social media, the sequencing of the human genome, and artificial intelligence, just to name a few. We are, without question, living in a brave new world.
As we accelerate into the increasingly technological era of the human story, we cannot afford to lose any more soul or spirit. We must preserve our humanity and what it means to be human as best we can for as long as we can. Being alive is the greatest gift of all, and being as conscious and aware as we humans are is an absolute miracle – perhaps the only miracle like it in the universe. Our planet is a better place when we recognize that each of our expressive and creative artifacts embody the spirit and gusto of what it is to be alive.