If art is the hallmark of civilizations, how will it evolve in the age of AI?

“Art is the signature of civilizations.” Jean Sibelius

Art is the signature of civilizations. The prominent Finnish composer Sibelius stated this. How will this signature be created? What form will it take in the era of artificial intelligence? What about art and AI?

Art has been a subject of philosophy for centuries as a way to grasp the world and humanity. Even defining art itself remains an endless topic of debate.

For instance, Plato viewed art as ‘mimesis’ (imitation). He defined a good piece of art by how well it imitates reality. Romantics saw art as the expression of emotions. Kant argued that art should be evaluated based on its formal qualities. This is in contrast to aesthetic beauty.

Corporate art theory suggests that for an object to enter the ‘art world’ context, it must be considered art. Examples include being displayed in an art gallery or museum.

—

The history of art dates back approximately 40,000 years. Art has played a central role in human life from the simple cave drawings of Neanderthals to the present day. It even preceded modern civilizations. Therefore, the definition of art undergoes constant debate and transformation from generation to generation.

Despite numerous theories and definitions, it seems that instead of becoming easily definable, art has become conceptually more ambiguous.

Today, this debate is being taken to another level with technology. Artificial intelligence has already become a new and powerful player in this discussion.

Although artificial intelligence is relatively new, the relationship between art and technology is not. The history of art produced with technological support dates back 50 years.

One of the earliest examples is British artist Harold Cohen. Cohen wrote a computer program named AARON in 1973 that created drawings by following a series of rule sets. While early versions of AARON produced abstract drawings, more representational images and color usage were added in the 1980s. Cohen later developed a method for AARON to paint using selected brushes and paints.

Technology has advanced rapidly since then. The space and abilities occupied within art have moved to a very different point. In fact, technology itself has become an art producer.

_{^{The First Artificial Intelligence Artwork: Portrait of Edmond Belamy}}

In 2018, the first artwork created by artificial intelligence was sold for $432,500 at Christie’s auction house.

A Paris-based collective art group named Obvius used an algorithm called Generative Adversarial Networks (GAN). This algorithm enabled artificial intelligence to produce the artwork “Portrait of Edmond Belamy.” Approximately fifteen thousand portrait images were loaded into the database. These images were produced between the 14th and 20th centuries for this purpose.

Developed in 2014, the Generative Adversarial Networks (GAN) model consists of two competing artificial neural networks: the generator and the discriminator. While the generator network produces new data resembling reality, the discriminator network tries to distinguish between fake and real data. As these two neural networks compete, the discriminator network gradually becomes better at distinguishing between real and fake images. Meanwhile, the generator network produces increasingly realistic fake images. If you feed thousands of cat images into the system, the model will learn how a cat should look. It will gradually generate more sophisticated new cat photos.

One of the artists who utilizes this model very effectively is Refik Anadol. Anadol’s ‘Unsupervised’ project elevates the GAN aesthetic and logic to the next level. It also opens up interesting visual avenues and conceptual avenues. For the project, Anadol trained the model with the MoMa’s 200-year-old collection database using machine learning. The result was fascinating image sequences displayed on a large vertical screen, pointing to the data source and allowing the reinterpretation of the works by establishing new connections among the original pieces.

Refik Anadol, who describes his work as a collaboration between human and machine, discusses the role of the artist and the creative process. He states, “With the same data, we can produce infinite versions of the same sculpture. Choosing that moment and creating that moment in time and space is the moment of creation.”

_{^{Artist Robots}}

The world’s first realistic artist robot is Ai-Da. Created in 2019, Ai-Da can draw and paint using cameras in her eyes, artificial intelligence algorithms, and a robotic arm.

Ai-Da’s first solo exhibition opened at the University of Oxford under the title ‘Unsecured Futures.’ Her paintings were exhibited at the Venice Biennale in 2022. Ai-Da continues to produce new works that deepen discussions on art and technology.

There is a heated debate in art circles. This debate is about works produced by advanced technologies such as artificial intelligence and machine learning. Opinions are divided into two extremes. One group considers this as part of art, while the other group insists that what emerges is not art.

My view is closer to the second group. Here are my reasons:

Art is emotion. However, artificial intelligence lacks the emotions that nourish art. It only operates based on commands. It cannot engage in a dialogue with values such as emotion, culture, symbol, language, belief, or period.

At least in its current state, it cannot generate new ideas. The content it produces relies on existing data. To create a new work of art, it imitates existing artworks. (ethical and copyright issues also come into play here!)

As a result, while algorithms may successfully mimic some human characteristics, can they help the audience understand life? Can they lead to a meaningful interpretation? I am not sure.

In closing, let’s leave the final words to Picasso, as this debate warrants further discussion:

“What matters is not what the artist does, but what he is. Even if his apples are ten times more beautiful, I would not be interested if he had lived and thought like Cézanne or Jacques Emile Blanche. What compels us to take an interest in him is Cézanne’s concern; that is the lesson Cézanne has given us.”

---