For decades, we’ve been taught that artificial intelligence will steal our jobs. It was when Kurt Vonnegut wrote the Player Piano when he shared that the workers of the world have been replaced by robots and machines. These ideas have recently found their way from the pages of books into the economic forecasting papers of governments and consulting firms.
While no one knows when or how many jobs will be lost to robots, it is commonly thought that garbage collectors, bus drivers, and interstate truck drivers will be among the first to go. However, it seems that people like me—creatives—are now much more vulnerable. New technology is making it clear that writers, illustrators, photographers, journalists, and novelists may soon be replaced by high-tech player pianos.
You don’t believe me? Take a look at the opening few lines of this short story:
The wind howls around you. It sounds like the voice of a god, I hear my raspy breathing and the sound of my tears hitting my lap and the floor below. The wind and trees croaked and creaked, a symphony of shipwreck, tumbling waves, and tortured life and death, as if the night sky were falling to the earth. The sound of my tears were heavy, plopping on the ground, and the sight of me.
Nobody with an MFA or who writes science fiction wrote that paragraph. Instead, it was put together by Sudowrite, a new online machine-learning platform that is marketed as a tool to help people write creatively. I wrote the first two sentences of that paragraph about the wind and tears, and the A.I. wrote the rest.
Writing progress is just the beginning. One of the latest AI programs to impress Silicon Valley residents is the Dall-E visual platform. It can take over any drawing and make it look completely different. The platform uses a variant of GPT-3, which learned to draw and paint by studying over a billion images. You can instruct it to make different kinds of art, or manipulate uploaded photos or images with amazing accuracy. Simply type a series of instructions into Dall-E, and it will quickly generate a graphic to show them. If you request a “photorealistic image of an astronaut riding a horse,” it will quickly generate one. If you tell it to do a “pencil drawing,” it will paint new images in that style, even making up the background. Stained glass, spray paint, Play-Doh, cave drawings, and Monet paintings are all options. Furthermore, you may replace the astronaut with almost anything else—a teddy bear, a dog, even Elon Musk—and the horse may ride anything else—another horse, a turtle, or “Millennial.”
These new technologies have upended our preconceived notions about creativity and computers. Philosophers have long thought that since robots lack sentiments, computers would never be able to make “art.” They don’t feel pain or pleasure, thus they can’t express such feelings artistically. However, it has been found that computers do not need feelings to produce art. They can only duplicate what humans have already produced.
But the more I see of these new machine-learning algorithms, the more I realize that the future is near. And in that future, many people may lose their jobs, not to mention their comprehension of what began in a person’s or a computer’s head. Soon, every time you read an article, you’ll be wondering if it was done by a human or an algorithm. We’ll never know for sure.