Can AI be truly creative?
Can humans?
Since ChatGPT was sprung on the world last fall, chin-tuggers have pondered whether it augurs the apocalypse or a new Eden. Will it turn us all into paperclips or find the drugs and proteins to cure every disease?
Persons less prone to extremes ask whether white-collar jobs will be decimated and plagiarism normalized. Will originality be lost when novels and screenplays, poems and music, paintings and sculpture become pastiches of previous works? Will genuine invention come to an end? It goes almost without saying—though it is said over and over—that generative artificial intelligence, which produces results based on patterns its software finds in existing materials, is essentially derivative. Artificial intelligence rips off the work of human intelligence. It can seem clever but it isn’t creative.
This assertion may be right, but perhaps for the wrong reason. The assumption is that humans are the creative ones, the only intelligences that can come up with new ideas. A computer can’t be Aristotle or Shakespeare or Leonardo or Newton.
It’s often said of artificial intelligence—and this is what gives rise to the world-ending worries—that nobody knows quite how AI produces its results. It infers its own rules as it goes. What ought to be said, more often than it occasionally is, is that no one knows how human intelligence produces its results either. This is particularly true of creativity. Where do new ideas come from? How new are they really?
Newton remarked that if he saw farther than others, it was because he stood on the shoulders of giants. He was being modest but not untruthful. He also said, when he was older, that as a young man he could hold concepts in his mind for long periods of time. John Maynard Keynes, no mean thinker himself, observed of Newton: “His peculiar gift was the power of holding continuously in his mind a purely mental problem until he had seen straight through it. I fancy his preeminence is due to his muscles of intuition being the strongest and most enduring with which a man has ever been gifted.”
Strikingly, what Keynes was describing is exactly what computers are good at. They don’t get tired; they can hold things in memory until their operators pull the plug. Newton would try one approach and then another and another until he found the path to the solution. He would have made the perfect computer.
Some geniuses seem to do this subconsciously. August Kekulé, a German chemist, came upon the hexagonal structure of benzene in a daydream. His mind was wandering and it drifted upon an image of a snake eating its own tail. Until then he couldn't get the carbon atoms to fit together in a manner that matched the data. Closing the loop did the trick.
He too was doing something computers are good at. He was familiar with the circular snake image from alchemy and other esoteric studies. His subconscious mind apparently took that and tried it out on his carbon studies. Et voilà!
French mathematician Henri Poincaré had been wrestling with a problem for months. He knew he needed a new idea to crack the puzzler. But the idea simply would not come. He decided to take a break and go on a geological expedition. As he was heading for the staging point, he boarded a bus. "At the moment when I put my foot on the step the idea came to me, without anything in my former thoughts seeming to have paved the way for it," he recalled. The rest was a matter of writing up the details.
What is creativity? Are there any really new ideas? Or just old ideas reshuffled, recombined and reconfigured? "There is nothing new under the sun," said King Solomon or whoever wrote the book of Ecclesiastes. Historians of art and literature make a living finding precursors to Michelangelo and Hemingway and every other pathbreaker. Ideas don't come out of nowhere.
We humans have long flattered ourselves that we're unique in the animal world. We do things no other animals can do. The list gets smaller the more we learn about those other animals. Crows can count; chimpanzees use tools; whales have language; elephants learn to recognize themselves in a mirror. But creativity—we've still got that to ourselves.
Yet if our creativity isn’t all we thought it was—if it's chiefly a matter of picking one shiny pebble from this pile, a second from another, and a third from yet another, and rearranging them—then what's so special about that? And what's so special about us? And what if machines can do it better than we can?
What Copernicus started, by decentering humanity’s home amid the cosmos, continues. Alas!
Or maybe that should be a sigh of relief. Being responsible for the universe is tiring.
Mulling over in my mind this morning this latest offering from Prof. Brands, “Can AI ever be truly creative? Can humans?” I am struck by the following: Last night, after driving for 11 hours, sitting with a beer in hand, Game 7 of the National League Championship on television, I opened and read Prof. Brands’ recent “Ever Victoria! Part 25.” When I came across the line: “The board calls Lib Tilton; she lies for Beecher now as she has lain with him in the past. (‘You always find the adulteress with the adulterer,’ one cynical lawyer observes.),” I stopped and re-read it several times, because it was a pleasing word play, and I truly appreciate fine word-crafting by word crafters in what I read.
This morning’s essay sparked me to think that last night’s polysemy appealed to me on an emotional level because the use of double entendre appeals to our sense of humor in the same way puns do. And I could not help but think I did not believe I had before read or heard this juxtaposition of forms of the words “lie” and “lain” making this novel, thereby adding to my enjoyment.
Then, I started remembering two philosophy professors whose lectures and personal conversations surely influenced my conscious mind: Prof. Irwin Lieb and Prof. Robert Solomon. Prof. Solomon stressed the role of our emotions, noting that it is our passions, not our reason, which are capable of demonstrating to us the meaning and purpose of life. Prof. Lieb once told me that one has to already know 80% of the answer to ask the question.
I have later come to realize that Prof. Lieb may have come to that sage advice having read it from an older sage. But, to me, I always associate it with him.
I then wondered if buried somewhere in my subconscious was an earlier exposure to the notion that to protect our mind’s personal narrative of self we sometimes recall as facts things that were really not facts, such as telling lies about those with whom we have previously lain. Lo and behold, I found that Ernest Hemingway in “The Sun Also Rises” used something similar in describing the relationship between Brett Ashley and Jake Barnes. And Mike Nichols used it in the film “The Graduate” to describe the relationship between Benjamin Braddock and Mrs. Robinson. Both of which, somewhere in my amygdala, hippocampus, prefrontal cortex or reticular formation, or maybe in all of them, are neuron projections and reinforced synaptic gaps of both of these examples which last night jarred an emotional memory recalcitrant to extinction when I read and reread Lib Tilton’s line.
To have consciousness, the subjective state of awareness of self and the world, we have to have a neocampus. And in the end, it is all chemistry.
So to the extent that the carbon and other elements that make up us and artificially intelligent machines all come from the stars and miraculously (or serendipitously, if you prefer) were assembled during the course of 4.6 billion years on our blue planet Earth into what became us, and out of us these machines, the answer to Prof. Brands question may be amenable to knowing, but it may take us another eon of millenniums to find out.