Creativity will never be efficient—and that’s a good thing

We are at a turning point. Our shock and surprise when ChatGPT went public two years ago has given way to the realization that it will be a long time before AI permeates our lives. Some of us are impatient, anticipating medical breakthroughs and increased productivity. Others fear—for our jobs, the most urgent—but also for the loss of human qualities that until now seemed untouched.

In Ernest Hemingway’s A Moveable Feast, a memoir about his time as a young writer living in Paris, he describes a daily habit of going to the museum himself after working on his book:

“In the afternoon I could walk through the gardens and then go to the Musee du Luxembourg where there were the great paintings that have now been moved mainly to the Louvre and the Jeu de Paume. I went there almost every day for the Cézannes and to see the Manets and the Monets and the other Impressionists I had first met at the Art Institute in Chicago. I learned something from the Cézanne painting that made writing simple true sentences not enough to make the stories have the dimensions I was trying to put into them. I learned a great deal from him, but I was not lucid enough to explain it to anyone. After all, it was a secret.”

This is a remarkable quote—a window into the evolution of a genius—but it’s also a warning about the dangers of optimizing creativity.

Here, a young Hemingway explores his art, tentatively feeling his way toward a new style. Day after day, he looks at Cézanne’s brushstrokes—the image of a mountain emerging from tiles of color, a complex whole made up of simple and similar parts. Day after day, he observes the same subjects from rocks and trees. They are in many of the tables, but always fresh. Like his sentences, this is the art that feels real and honest and true to him.

There is nothing surprising when one in the 1920s is confronted with Cézanne’s innovation. The painter was then modern, even radical. The unexpected thing is that Hemingway turned to Cézanne as inspiration for his writing, discovering a path from color and design to narrative structure and the rhythm of words.

It worked. Hemingway’s style profoundly shaped 20th-century prose and influenced every writer who followed him. But the way he got there was not obvious or efficient.

LLMs—Large Language Models—the AI ​​engines behind chatbots like Google Gemini and Open AI’s ChatGPT, work by computing the most statistically plausible next word in their training data, and then the next most plausible word after that next. For many purposes, this AI-generated content is enough—it’s conventional, it fits the job, and it’s incredibly fast to produce. But an LLM’s aim to deliver the best result means that what we get is formal and often derivative.

That Hemingway improved his writing style by looking at a Cézanne painting was anything but typical. It would be more efficient for him to find an example in another writer. But Hemingway was doing something new, not looking back and optimizing his creativity. He went to the Musee du Luxembourg and stood in front of a painting, until a new way of writing appeared in his mind — so vague, at first, that he could not even tell what it was.

Throughout history, the most original thinkers and creators have often found inspiration outside their core disciplines. Perhaps multimodal artificial intelligencemachine learning models that work with many kinds of inputs will reflect this feature. The results from these models are, in fact, are starting to appearespecially in scientific research. Or maybe experiment with chain of thought (CoT) prompt. it will introduce human reason. OpenAI-o1, was released on September 12 this year is the first in a new series of CoT models designed to “…spend more time thinking before responding”.

In some circles, there is anticipation that AGI (artificial general intelligence) or ASI (artificial superintelligence) will incorporate knowledge from any number of disciplines—perhaps *every* discipline. But even this may not be the key to elite human-level creativity.

Hemingway was curious—and patient. He understood that new ideas emerge from unexpected places and take time to coalesce. He accepted that human thinkers get distracted, take wrong turns, and are troubled by doubt. He worked hard at his art, but welcomed daydreaming and wandering the gardens.

The GPS in our phones may have already claimed the ancient skill of navigation. It’s been at least two generations since electronic calculators diminished our desire and ability to play with numbers. And without regular practice, human traits like imagination, observation, and curiosity will rust and fall away as well.

As we increasingly turn to genetic artificial intelligence to help us create—the trait that separates our species from all others on this planet—we must ask ourselves whether an “optimal” process is really what we want. Especially from a tool that is supposed to help.