Advancements in technology have always impacted creativity. From Michelangelo crushing lapis lazuli for the Sistine Chapel to the development of the early iterations of Adobe Photoshop in the late 1980s, artists have created work in collaboration with the progressive tools at their disposal. Now, as we enter an age where AI can actually generate art as opposed to assisting humankind in making it, what does it mean for the future of our collective, creative output? Are we worried? Maybe a little. Should we be excited? Definitely. Writer and author James Bridle assesses the current landscape and looks at the actually-not-too-scary future.
ASCII Art by Joan G. Stark, Joris Bellenger and Michael J. Penick, commissioned by Turbo.
An unfortunate incident occurred at the Moscow Chess Open tournament in the summer of 2022. In the pithy phrasing of Sergey Lazarev, president of the Moscow Chess Federation: “A robot broke a child’s finger—this is, of course, bad.” The seven-year-old boy in question apparently made his follow-up move too quickly for his robot opponent’s liking, which grabbed and held him until he was freed by bystanders. He recovered quickly and played the rest of the tournament in a cast.
Chess has a long history as a field in which our hopes and fears for artificial intelligence have been tested against human ability and creativity, from the earliest chess programs of the 1960s, to the infamous 1997 face-off between Garry Kasparov and IBM’s Deep Blue, and beyond. One of the results of this is that we tend to see AI as an opponent: a competitor with, rather than a collaborator in, our own hopes and desires. It’s easy to see how narratives of superintelligent overlords flourish when the most prominent examples seem determined to thrash us in the very pursuits we engage in for pleasure—and profit.
Lazarev’s direct assessment of the situation—“This is, of course, bad”—is a phrase you don’t hear too often in the ongoing debates about the effects and ethics of emergent artificial intelligence, and it’s refreshing to hear it stated so plainly. Although the incident in question has little to do with AI, and more to do with the poor implementation of complex mechanics, it nevertheless struck a chord with the global media, who reported the event widely, and frequently in tones which hinted at some kind of robot uprising. The sensational trope of man versus machine has a long history, and constantly finds new ways to express itself in our technology-obsessed culture.
This oppositional stance towards the flowering of novel forms of intelligence is unfortunate, but hardly surprising, when so many of its manifestations are explicitly designed to outmaneuver and supplant human intelligence. When the World Chess Champion Mikhail Botvinnik developed some of the first advanced chess algorithms in the 1960s, he could look for inspiration as far back as 1770, and the debut of the Hungarian inventor Farkas Kempelen’s Mechanical Turk, which wowed audiences across Europe with its mechanical skills, before the eventual revelation that it was really a series of impoverished grandmasters locked in a box.
Early computer chess programs were mere shadows of the kind of complex systems now arrayed against game players of all kinds, from AlphaGo, the machine which defeated the Go professional Lee Sedol in 2016, to Pluribus, a “superhuman” poker-playing AI developed by Facebook and Carnegie Mellon, which was deemed so threatening to the game that its creators refused to release its source code to prevent others using it to cheat in online games.
This perceived threat to pleasure and profit is also what animates concerns over the potential creativity of artificial intelligence, embodied in systems as diverse as Ai-Da—the “robot painter” created by a team from Oxford University—whose work has been exhibited at the Victoria and Albert Museum and the 59th Venice Biennale, and DALL-E, the image generation software from OpenAI, which has spawned a thousand op-eds on “the end of creativity.” (Perhaps it’s notable that the former is named for Ada Lovelace, the first computer programmer, who noted in her memoirs that her machine, the Analytical Engine, “has no pretensions to originate anything. It can do whatever we know how to order it to perform,” while the latter was inspired by the endlessly quotable Salvador Dalí, who reportedly wrote that “Those who do not want to imitate anything, produce nothing,” and “What is important is to spread confusion, not eliminate it.” His namesake certainly seems to have taken these directives to heart.)
DALL-E is manifestly good at generating mostly—if somewhat eerily—convincing images in response to carefully-tuned prompts, in much the same manner as an overworked magazine illustrator in hock to an increasingly manic and micro-obsessed commissioning editor. As many have pointed out, we seem to be entering into a new era when “prompt engineering”—the skill of crafting the most succinctly tailored brief for the AI “creator”—becomes the dominant form of human creation.
On the one hand, there’s a legitimate concern here that humans are reduced to mere bot herders, mangling creative demands into a word salad that approximates the understanding of the machine (think “shallow depth of field thomas kinkade popular on artstation light upscale max” commands, ad infinitum). In our interactions with inhuman technologies, we all too often come to resemble them. Yet on the other hand, the concern over individual artistic agency and responsibility is hardly new. Creative works of all kinds have always emerged from a complex, entangled web of social, emotional, and financial demands, concerns, desires, and experiences. The great Renaissance painter Tiziano Vecelli, known to history as Titian, employed a studio of dozens of assistants, many of whom made whole careers there, and their anonymous hands are well-documented in many of the works considered “his” masterpieces. The famous Mühlberg portrait of Charles V shows evidence of at least seven of them, while many of his most celebrated works were completed by others after his death, his entire oeuvre a kind of prompt engineering for generations of European art.
If that seems a frivolous equivalence, then consider it this way: “creativity,” like intelligence, is not something which occurs in a vacuum, or in a single mind. Rather, it is an emergent property of encounters, experiences, materials, and relationships—only some of them identifiably “human,” and many of them non-human and entirely constructed. The luminescence of an Early Renaissance fresco by Fra Angelico owes as much to red cochineal bugs, tempera from hen’s eggs, the fur of his sable brushes, lapis lazuli from the mines of Afghanistan, and ecclesiastical fashion and historical taste as it does to the inspiration of the individual. Creativity has always been a social, interpersonal, and interspecies phenomenon. What we call human art is revolutionized over and over again by the intercession of non-human being and materiality, from oil paints to silver halide photography.
What may be significant about the present moment is the gnawing fear that the latest player to enter the game—machine intelligence, whatever that means and however it manifests—might somehow surpass our own abilities. What this fear betrays is our powerful, shameful sense of human superiority: that we create what we create because we are the smartest, most brilliant beings on this planet, and thus occupy some special, unique place in the universe. We fear the loss of this superiority when confronted by machines which beat us at chess, take our jobs, and make things which look a bit like art, or at least the kinds of art which many humans depend on for money. But it is this self-same sense of superiority which has led us to our present planetary crisis, and if nothing else, AI might be the spur we need to recognize it for what it is, and move beyond it.
Human intelligence is unique, but it is not more unique than any other kind of intelligence. Other primates have their own ways of doing things, from complex tool use to social strategies, that mirror their own particular circumstances. Honeybees use their intelligence to communicate the location of nectar-laden flowers, and vote democratically on new nesting sites. Slime molds—single-celled organisms which live on dead wood and forest floors—display extraordinary computational abilities, far in advance of both our own mathematics, and that of the most powerful supercomputers. The examples, when you look for them and want to see them, are endless.
These forms of intelligence arise as a matter of context and relationships, of individual and collective experience as they manifest in the world, and go to work with it. Intelligence flowers everywhere, and thus there is really no such thing as “artificial” intelligence, even in machines. What we call “artificial” intelligence is just another instance of this going on together, as much a part of the “natural” world as oceans and oak galls, mushrooms, mitochondria, and machines made from long-dead sea creatures. “Artificial” and human intelligence, and the creativity which flow from them, are not, and should not be perceived as being, in competition with one another, but are equally fascinating and valuable ways of exploring and making sense of the world.
What matters when it comes to intelligence and creativity is the manner and intention of our relationships. This is why I’ve always distrusted notions of AI-based on competition, those chess computers and poker programs, which model intelligence as a kind of all-out war of mind against mind. It’s like raising a kid exclusively on Machiavelli and the “Art of War:” of course they’re going to turn out as mean and probably quite successful little bastards, but they’ll be swiftly drummed out of any community they can’t despotize. Which is what they’re trying to do now—to take over playful and creative communities with the goal of turning them into profit-extraction industries—and why we should take the idea of community seriously if we want to live with them in more just and equitable ways.
The uses of AI that seem more interesting to me than rote prompt engineering and nuclear poker—and indeed, finger-biting robots—are those which engage meaningfully with the world around them, in ways that build rather than destroy relationships with natural systems and non-Western cosmologies. We’re already seeing how DeepMind, the company which produced the uber-competitive game engine AlphaGo, has turned its technology towards protein folding, one of the hardest problems in biology, with the promise of as-yet-unknown, but vital, medical benefits (although, once again, we should be wary of where the profits of such discoveries accrue). Meanwhile, groups such as Indigenous AI in North America are broadening the conversation around AI ethics to include the needs and desires of all humans and non-humans; the Māori technologists of Te Hiku Media in Aotearoa New Zealand are developing protocols for natural language programming, which respect the history and sovereignty of traditional knowledge keepers.
In 1997, following his infamous defeat by IBM’s Deep Blue, a computer developed explicitly to beat him, the world champion grandmaster Garry Kasparov convened the first chess tournament in which humans played alongside, rather than against, machines. It was quickly discovered that this form of play, in which computer programs and human players supported one another, opened up entirely new fields of gameplay: matches in which new styles and gambits quickly developed, and which outmoded in novel and expressive ways earlier forms of the game. Indeed, it turned out that in many instances the combination of human and computer program was more powerful than any machine acting alone: a testament to the power of co-operation over competition, which should give the lie to the oppositional thinking so ingrained in our relationships with AI.
One name for this type of play is “centaur” chess: the melding of human forms of creativity and ingenuity with radically different forms of intelligence: mythological, non-human, and machinic. It turns out that co-operation, and the recognition of ways of thinking and being beyond the human, is the key to our mutual flourishing and development. Relationships formed in respectful communities endlessly engender new perspectives, and AI is now part of our more-than-human community. “The end of creativity?” Never. Only, once again, a new beginning.