Arts, Column

Tran: Why Artificial Intelligence Will Improve the Existence of Art

An empty text box and the bolded word “Generate” make up the minimalist user interface of DALL-E, a deep learning engine created by OpenAI, the same developers of ChatGPT. With a simple command, DALL-E can generate dozens of variations of any requested artwork in seconds. 

As I set out to write this column, I made DALL-E present me with these three images: 

Before hitting “Generate,” these three exact images did not exist—not even in my mind. DALL-E took the few words I wrote in the prompt—such as “80s style Polaroid photo” and “four purple bears dancing around a fire with one playing an accordion”—and brought them into existence. 

Artificial intelligence uses code to source and bring to life combinations of images, art styles, and abstract ideas, typically generated by human intelligence. Similar to how different people have their own idea of what love, music, and emotion would look like visually, DALL-E can bring to life its own imagined pictures of these abstract concepts. The model can create the simplest of line drawings to complex photorealistic portraits. Limitations to what DALL-E is capable of are difficult to discover within the deep learning model. 

The implications behind AI’s growing abilities can be seen as disastrous. Among those affected by the growth of artificial intelligence are graphic designers, small freelance artists, and book illustrators. While businesses can save the cost of hiring an artist to create logos and event posters with AI, the new landscape of art puts many out of work and others in fear of a total AI art takeover. 

Although critics have reassured readers that AI art will not threaten the art industry, their reasoning boils down to the idea that human art is irreplaceable, and that AI art will pass as a fad because of AI programs’ limitless ability to create art. Because AI art is so easily created, none of it is valuable.

I, however, believe DALL-E is here to stay, and it and other art-generating AI programs will only better the trajectory of creating art. 

Frankly, after decades of remorse and controversy over contemporary art, AI art may be the push the art world needs to steer artists toward new sights that DALL-E can not replicate. 

While DALL-E mainly creates tired digital art fit for Etsy throw pillows and social justice infographics on Instagram, its existence will push artists to finally enchant onlookers with more nonconformist creations. This leaves engines like DALL-E to eternally play catch up to artists’ imaginations, and moreover, push art away from a stylistic standstill. 

New artificial intelligence is able to mock contemporary art quicker than a critic with a paintbrush. Once it realizes the ability of DALL-E to replicate art, the industry will not be able to hide behind mere abstracted meaning, forcing artists to look to push themselves stylistically. 

When art suprematist and historian Camille Paglia originally hailed secularism as the cause of the laziness of contemporary art, AI art was an imaginary threat. In Glittering Images, she argued that there is a new scarcity of artists like Andy Warhol, who employed his Eastern-Rite Catholic roots to create a Byzantine-like art for the modern pop world. Pop art was sparedfrom  being criticized as shallow because of driving contextual passion and inspiration. Today, passion that is derived from ideas like spirituality is purged from contemporary art.  

Her reaction today may parallel the idea of AI being the final realization to the art world’s fear of disrepute. Where complacency has plagued art for decades, perhaps the introduction of AI may lead to a dramatic shift.  

Returning to the three images I had DALL-E create at the start of the column, there is one qualifier to be made about how the AI engine makes the art. 

The text box displays an example prompt that someone can type for DALL-E to generate. 

Courtesy of OpenAI

The example suggests to the user to be specific in the style of art that they want the engine to create, such as replicating an “Impressionist oil painting.” Doing so lays the central misunderstanding of OpenAI’s purpose and capability for DALL-E. Instead of taking over art with its own creations, DALL-E can only mimic human creativity. 

But only to an extent. 

To put DALL-E to the test, I told it to make a fourth image to mimic my favorite artist’s style. Ernst Kirchner dominated expressionist art in the early 20th century with broad stroke depictions of extravagant German life, replete with a cool palette and eerie atmosphere. It’s safe to say what DALL-E produced looks nothing like it. 

DALL-E proves its use in helping those curious enough to see how far their imagination can go in words. The hard part when using a tool like this is when you already have the visual idea in your head. The technology ends up being disappointing when it fails to create the exact image you had imagined while typing the prompt.  

Like ChatGPT, the use of AI generated art can be a double-edged sword. In both programs, the potential for moving society away from contentment and complacency is clear. The use of AI will create a necessity for aspiring writers and artists to move away from falling into the same repetitive or uninspired work. The saying “if it isn’t broken, don’t fix it,” applies here, as technology has come up with a way to finally break the tired pattern of art. This shift will be inevitable as AI becomes fixtures in our technologically advancing world. 

I predict and hope for a new movement that will arise from the introduction of AI art. What that style of art will look like will at first be unrecognizable to us—that is the highest sign of creativity. 

February 26, 2023