So it shouldn’t have been so startling to see the technology write back to us. Yet these artificially intelligent answers, so clear and knowledgeable, seemed to signal something new, something deeper. Even in the initial rollout, AI seemed to have leapfrogged three-quarters of human beings at this distinctively human endeavor. Textbook writers and a certain sector of nonfiction seemed doomed. The technology seemed to have shown up almost fully formed, effortlessly lucid on most topics, serene, almost smug in its mastery of journalistic prose. Prompts beginning “Write a poem…” proved that it could turn a rhymed quatrain more reliably than many practicing contemporary poets, too. Within a few months, we were experimenting on text-to-image generators, astonished at how our words could transform, within seconds, into Persian miniatures, or paintings in the style of Goya. It wasn’t half bad. In fact, it was quite good. What would old-fashioned artists do? The advent of photography had shot realism in the stomach, but AI seemed likely to shoot art itself in the head. And then came text-to-video generators, conjuring uncannily sophisticated clips from motionless, colorless words. Did this all really happen in a year?

To read the full essay, please download the PDF at the link below.