Has the AI-Art Movement Arrived?

AI painting “La Comtesse de Belamy" | Image source: vip.gatech.edu
AI painting “La Comtesse de Belamy" | Image source: vip.gatech.edu

We know that there is a myriad of business applications for artificial intelligence (AI) and machine learning – from conversational interfaces to predicting software vulnerabilities to automation. We use AI more often than we realize. After all, we interact with our smartphones, access social media, use Google Maps, or call an Uber – all of which use the technology in some form. With AI already integrated into our everyday lives and society’s acceptance of it, it only seemed a matter of time before people found further applications of the technology. And as predicted, AI emerged in the art world, intriguing many and angering others.

The creation process of  “La Comtesse de Belamy,” an artwork created by a Generative Adversarial Network (GAN) by OBVIOUS, the three men group of French art collective.

Is AI positioned to be art’s new medium?

Art has a rich history of collaborations. Artists working with other artists have resulted in some brilliant works of art. From 1980 to 1986, Andy Warhol and Jean-Michel Basquiat interacted and cooperated on pieces such as the Olympic Rings. Together, Pablo Picasso and “lighting innovator” Gjon Mili work together on the exquisite Drawing with Pure Light.

However, artwork that is the product of the minds and skills of two phenomenal artists is one thing; what about art created between a human and machine? Now, what about art created by a machine with artificial intelligence? Because while AI is a human invention and runs on an algorithm defined by a person, the result it produces has a limited human intervention – making it an artistic collaboration that is both compelling and questionable.

AI art is artwork created by an algorithm defined by an algebraic formula and machine learning technology. Artists set the rules or code, and the algorithm analyzes thousands of images to “learn” specific principles of art. The algorithm then generates images by imitating the aesthetics it’s learned. The only real interaction the human artist has with the process is choosing what pictures to feed the algorithm. The artist may also tweak the algorithm to get the results they want. Because the machine studies thousands of images, it typically produces some rather bizarre results – from distorted faces to deformed bodies to unrecognizable landscapes. For this reason, the AI art has fallen into the conceptual art category.

Pierre Fautrel of Obvious poses with the artwork, which sold at Christies for $430,000 in October 2018 (Credit: Getty) | Image source: bbc.com

Setting the stage for the AI art movement

The debate on whether AI art is “real art” is ongoing. And whether you strongly believe that anything AI-generated doesn’t belong within the art world’s parameters, it’s clear that the AI art movement isn’t just the future, it’s here, and the stage for AI art has already set. Last year, Portrait of Edmond de Belamy (2018) made it to auction, proving that the AI art movement has arrived and that there’s a market for it. Christie’s estimated the piece at only $7,000 -$10,000 but ultimately, sold for an astonishing $432,500 – signaling what Christie’s describes as “the arrival of AI art on the world auction stage.”

Portrait of Edmond de Belamy (2018) is the product of GAN (Generative Adversarial Network), a class of AI algorithms utilized in unsupervised machine learning. It’s used to generate photorealistic images and reconstruct 3D models based on pictures.

Behind Portrait of Edmond de Belamy (2018) are three men who form the French art collective, “Obvious.” Its creation involved feeding a collection of 15,000, 14th to 20th-century portraits into a machine learning system. Tools called the “Generator,” and the “Discriminator” created the image based on the data they entered into the system. Obvious collective member Hugo Caselles-Dupré describes the process, “The Generator makes a new image based on the set, then the Discriminator tries to spot the difference between a human-made image and one created by the Generator. The aim is to fool the Discriminator into thinking that the new images are real-life portraits. Then we have a result.”

Nude art images by AI artist Robbie Barrat using progressive growing of GANs (generative adversarial network) | Image source: independent.ie

Emerging AI artists to look out for

Thanks to savvy marketing team of Portrait of Edmond de Belamy (2018), enough buzz and controversy surrounded it to get it the attention it needed for a lucrative payout. However, Portrait of Edmond de Belamy (2018) isn’t the first AI-generated artwork to make it to auction and the collective, Obvious, aren’t the first to use AI in their creations.

Anna Ridler is an artist who has worked with AI for a few years now. While Ridler holds traditional art degrees from the University of the Arts London and the Royal College of Art, she has programming skills to generate datasets. She also typically uses GANs in her work, describing working with GAN as a way for her to discover new possibilities. She explains, “It has held up a mirror to my own process and allowed me to see elements that I was unaware of (how I draw eyes, what I choose to emphasize).”

Digital artist and MIT-trained roboticist Alexander Reben describes the creative potential of AI as “endlessly interesting.” He tells The Washington Post. “It’s the ultimate artist dream to lie in a hammock and think about art that is then generated for them.” Reben’s artwork has already been shown internationally, exhibiting in venues such as Ars Electronica, the Boston Cyberarts Gallery, and ArtBots.

Robbie Barrat describes himself as an “Artist working with artificial intelligence.” The art collective, Obvious, admits that the algorithm they used to create Portrait of Edmond de Belamy (2018) is largely based on 19-year-old Barrat’s algorithm – a code that he shared via an open source license. And while others benefited from his programming, Barrat is still making quite the name for himself in the AI art community, particularly for his landscapes and nudes. He even continues to share his algorithms to help aspiring AI artists and elevate the AI art movement.

OBVIOUS signed their art, a print on canvas, “Portrait of Edmond de Belamy” with the algorithm used to create it. (Credit: Getty) | Image source: bbc.com

From the fringe to the high end

The AI art movement has arrived, and with it, the rise of AI artists who are ready to defend their work and collectors prepared to pay thousands of dollars to possess a piece of the movement even if its still in its infancy. While art critics and traditionalists may continue to question the AI-generated imagery place in the art world, many argue that it deserves its moment and respect just as much as Duchamp’s “Fountain” – the urinal that has forever changed how we define art.


Leave us your question.