SINGAPORE — At the heart of Ken Liu’s quiet, deliberate voice lies a paradox: copying is part of how art is made.

Liu does not mean copying word for word or plagiarizing. What he’s talking about is imitation as learning and transformation, not duplication.

Speaking before a captivated audience at the Singapore Writers Festival, the award-winning author and futurist dismantled one of the most persistent myths in Western aesthetics — that art is defined by originality. 

“In the West,” he said, “we traditionally think of art as a matter of creation, of doing something new. But that has not always been the case. The foundational aesthetic theory was that art is about copying.”

From Aristotle’s idea of mimesis to Sir Walter Scott’s praise of Jane Austen for “copying from nature,” Liu reminded listeners that imitation, not invention, was once the truest sign of artistic mastery. His talk, which moved fluidly between history, philosophy, and technology, explored what happens when machines begin to copy as well.

When machines learn to copy

“What distinguishes human copying from machine copying?” Liu asked. 

“Over time, machines have taken over more and more of the copying function. By studying that process, we can understand why certain arts are threatened by mechanization — and why others are not.”

To illustrate, he led the audience into the medieval scriptorium, where scribes once copied books by hand. Long before printing presses, bookmaking was a slow, creative ritual. A scribe was not just a copyist, Liu explained, but an interpreter — one who added meaning with every stroke of ink. 

“Copying a book by hand was a creative exercise,” he said. “Each scribe decided how to divide words, how to interpret sentences. Reading and writing were both acts of art.”

When the printing press arrived, this tradition vanished. The machine could reproduce a single interpretation millions of times — a miracle of consistency, but at the cost of human variation. 

“We eliminated that art,” Liu said. “We decided that interpretation at the level of reproduction was no longer worth preserving.”

Art that machines replace

He then shifted centuries forward to the birth of photography. 

Painters, Liu noted, were not the only ones unsettled by the camera. Engravers, artists who translated paintings into prints, were the ones truly displaced. 

“An engraver was like a translator,” Liu said. “She had to interpret a painting into another medium.”

When photography arrived, artists could reproduce their own work without relying on another creative intermediary. The camera, unlike the engraver, had no ego, no interpretation. It was mechanical fidelity incarnate. 

“Even though a photograph is an interpretation,” Liu explained, “it is an interpretation within the author’s control. The reproduction has no additional creativity added to it.”

In this way, photography performed the same cultural surgery as the printing press, it replaced a human act of interpretation with a mechanical process of replication. “And within a generation,” Liu added, “engravers were gone.”

Copying, cheaply

Liu lingered on the irony of the word image itself. 

Derived from imago — the final stage of an insect’s metamorphosis — it originally meant a “perfect copy.” Humanity, he mused, now lives in an age dominated by images: copies without originals, reproductions without craft. 

“It is actually very rare to see an original anymore,” he said. “Almost everything we see is a copy. And we like it. We love living in a world inundated by cheap sloth.”

But rather than despair, Liu offered a kind of invitation: to look more closely at what makes human art human. The difference, he argued, lies not in the result, but in the process — in the slow, deliberate craft that imbues meaning over time. 

“Art is about growth,” he said. “Mechanistic copying is instantaneous. Art is about process leading to a result; the machine has only the result.”

Human in the Loop

What happens when AI can write, paint, and compose as well — or better — than us?

Liu did not offer predictions, only possibilities. One future may see robots producing novels and films faster and cheaper than any human can. Another may reject machine-made art entirely. Most likely, he suggested, we will live somewhere in between — still unsure which parts of ourselves to surrender to the machine.

“If an art’s highest ideal is to reproduce an external standard of originality, it is subject to mechanization. But human art — art that grows, that interprets, that crafts meaning over time — will always resist.”

Liu’s vision of art is not one of fear, but of reckoning. As machines learn to copy, perhaps what matters most is not the originality of what we make, but the humanity in how we make it.