The Evolving Role of AI in Art — Wondrous Yet Profoundly Unsettling

Salvador Dalí was one of the greatest painters of the 20th century, revered for his creation of vivid, surreal images of melting clocks and elephants with incredibly long legs. Known for his flamboyant personality and sharp mustache, Dali would take up to a year to finish one of his unique and thought-provoking works of art.

In 2023, you can create surreal images in the same style as the great master in a matter of seconds, using an AI program honoring his name — DALL-E 2. To the untrained eye, some of these images may even look like actual paints created by Dali.

Naturally, not everyone is pleased by the increasing role played by AI in the world of art — some may even call it an intrusion that threatens the future of human artists. Others champion it as the future of art. The truth as often happens in these debates lies somewhere in between.

Making a Distinction Between Computer Art and AI-Generated Art

Using computers to create new forms of digital art is nothing new. One of the earliest documented examples of computer art is from 1963 — the cover of Computers and Automation weekly depicted a Black and red waveform art created by a junior technician at MIT called Ebram Arazi.

These early images relied heavily on geometric shapes, lines, and patterns in monochrome. Since then, artists and graphic designers have continued to use computer software as both the tools and canvas to create increasingly complex and colorful artwork.

Modern digital artists use computers and tablets along with software like Adobe Photoshop and Corel Painter to create artwork. The process takes time, considerable skill, and hand-eye coordination, not to mention a lot of imagination.

AI-generated art does not require any of this. All you need to punch in some keywords and the software will generate unique new images that closely approximate those ideas in a matter of seconds. For instance, to get a series of Dali-esque images, all I had to tell DALL-E was “a surrealist painting of animals in the style of Salvador Dalí”.

The process does not require a lifetime of skill, training, or practice. This is one of the main things that separate AI-generated art from art created by digital artists.

Yes, Even AI Can Generate “Real” Art

In September 2022, a digital image “created” using the AI program Midjourney won the top prize at an art competition at the Colorado State Fair. The judges did not know that the image was created using an AI when they awarded it the top prize.

This raises an important issue about the validity of AI art. Many critics argue that images or designs created by AI at the behest of prompts by human users do not count as “real art.” They cite the lack of effort in the creation process as a major flaw.

But the merit of an artwork is not just based on the toils or skill levels of the artist — it is largely decided by the audience. As the great filmmaker Abbas Kiarostami once said: “A work of art doesn’t exist outside the perception of the audience.”

If an AI-generated image is capable of impressing human viewers and generating an emotional response, the artwork does gain significant credibility, there can be no doubt about that.

Empowering Tool vs Exploitative Existential Threat

Many artists are rightfully afraid of the threat posed by AI tools like DALL-E, Midjourney, and Craiyon. There are immense commercial possibilities for AI-generated art, in marketing and other creative-oriented fields.

Software capable of generating dozens or even hundreds of unique images in a matter of seconds, tailor-made to user prompts and specifications, could replace human artists in many fields in the coming years.

And perhaps the ultimate insult — the AI are trained using images and artwork created by humans themselves. The more you create and publish — the more the AI can learn and improve — for instance, DALL-E-2 was trained using hundreds of millions of images.

Not all artists are hostile to the rise of AI. German artist Mario Klingemann is a pioneer in the field of AI-driven art. He was the driving force behind the creation of Botto, a community-driven algorithm that has generated over $5 million in sales from auctions of its artwork.

And it is not just in paintings that AI is making an impact — The Rain Room by the Random International art collective from 2012 is another stunning early example of the potential of AI algorithms in interactive experiences. Since then, art festivals that showcase the creative use of AI have become a regular feature of the international art scene.

And we also see the rise of artists who are also proficient in AI and robotics. Sculptors like Alexander Reben use algorithms to generate inspiration for new projects. Lozano Hemmer is creating interactive installations powered by AI that use visitors’ biometric data to create unique experiences.

AI Can Create Art, But They Are NOT Artists (For Now)

Visual AI tools like DALL-E and its literary counterpart ChatGPT still have a long way to go. In their present state, these algorithms are still mindless automatons with the ability to churn out original but often problematic content only at the behest of humans.

A true AI, or AI General, would be capable of creating artwork without relying on prompts, based on an innate urge (just like us humans). The software at present have got part of the artistic process right — they are drawing on experiences gained (from studying other images), analyzing the basic structures, and then creating something new altogether.

However, the creative spark, the driving force behind this process, is entirely human. The experience — or rather, the collective repository of a history of art, is also fully created by humans. Even now, AI is relegated to the role of a tool — albeit a tool with unprecedented creative powers.

There is no doubt that we need to be careful about the commercialization of AI-generated artwork, especially when it comes to the unauthorized use of source material, styles, and the potential for racial and cultural bias. At the same time, we have to accept that AI is here to stay in the art world, for better or worse. The outcomes depend on how we choose to utilize this powerful technology.