OF ART, ALGORITHMS AND ARTIFICIAL INTELLIGENCE
What will become of the artist as machines are learning to create art? Sonia Kolesnikov-jessop investigates
While German artist Mario Klingemann has been fascinated by artificial intelligence ( ai) ever since he read Marvin Minsky’s The Society of Mind in 1989, it wasn’t until recently that technology has matured enough to allow him to make his ideas a reality.
Today, Klingemann creates artworks using algorithms, code and neural networks. He is considered one of the art world’s pioneers in the use of computer learning and even won the 2018 Lumen Prize Gold Award – an international award that recognises the very best art created with technology – for The Butcher’s Son, a sitting nude image entirely generated by a machine using Generative Adversarial Networks ( gans): a class of ai algorithms used in unsupervised machine learning.
First developed in 2014, “these algorithms are called ‘adversarial’ because there are two sides to them,” explains Ahmed Elgammal, Director of the Art & Artificial Intelligence Lab at Rutgers University in New Jersey. “One generates random images, and the other judges them and deems which is the one that best aligns with the input.”
For The Butcher’s Son, Klingemann built a system with multiple neural networks that explore the human form. “It first generates random stick figures after having learned
the possible variations in human posture by analysing hundred thousands of photos. Each of these figures is then passed to a gan that has learned to translate it into a rough sketch – interestingly, that painterly look is the result of the model making mistakes and abstractions, since I did not train this model in painting; I actually trained it in photos,” Klingemann says. “In the final step, this sketch is given to another gan, which tries to fill in the missing details as well as add interesting textures and artefacts.”
The final winning portrait, with its disfigured face and blurred flesh- coloured tones, is reminiscent of Francis Bacon’s artworks, in line with a distinct gan aesthetic that reflects how the algorithms process information.
“Visually, there are typically surreal distortions in localised areas, which create a very interesting part-whole relation in the painting: what seems like a typical beach landscape might suddenly in one region sprout a forest,” says Karthik Kalyanaraman, one half of the curatorial collective known as 64/1, which focuses on art for the post-human age and curated a show last year entirely dedicated to ai art at Nature Morte, one of the leading contemporary art galleries in India.
Kalyanaraman points out that an art expert can usually identify work produced by a gan because of telltale textural effects, “a kind of shimmer or a scrambling of various image-making techniques such as impasto and dry paint”, though artists are now working on overcoming this
gan signature.
Elgammal, for example, has developed a system he calls aican: a creative rather than a generative network that is specifically programmed to produce novelty. It was essentially developed as a “creative collaborator and
partner” for artists, and some of the works produced using the aican in “collaboration” with artists Tim Bengel and Devin Gharakhanian were premiered at scope Miami Beach in December last year.
But ai- generated art had piqued curiosity even before that; it was put in the spotlight with a controversial sale at Christie’s New York in October 2018 that saw a piece of ai- created artwork smash its US$7,000 to US$10,000 pre-sale estimate to eventually sell for US$432,500. The piece in question? Portrait of Edmond Belamy, created by a
gan trained by Obvious, a Paris-based arts collective that signed the work with the algorithm’s name instead of its own, which might be seen as an acknowledgement that the collective considers the algorithm the artist.
So... who is the artist? Is machine intelligence about to replace humans? Not so, say art practitioners.
“The narrative that the ai is the ‘artist’ is absurd and that is clear once one starts to understand these algorithms,” says Kalyanaraman. “Humans conceive the algorithm, teach it a particular visual style by curating the ‘training set’ and use their aesthetic eye to curate the final output! In so far as there is not a shred of autonomy or will (on the part of the ai) in this process, I think it’s really premature to call the ai an artist. Do we consider the lens and the camera and ‘nature’, which provides the setting, the true artist and not the photographer?”
Klingemann adds: “A gan by itself is often just an empty vessel. As an artist, you have to fill it with content by deciding what to train it in, and then finding and curating thousands of images that you want to extract some essence from.”
He also points out that “machines can probably only ever be as creative as humans, not much more. Even if they came up with radically creative ideas, we would not be able to recognise their brilliance, since we would simply not be able to understand it. That’s the problem with human imagination – it can only expand slightly beyond our (current) horizon; new ideas take time to be accepted and understood.”
In what might be seen as a slightly Orwellian way, he remarks: “Like any other technology that allows us to do things faster, better or with less effort, ai already seeps into our daily life, and will not stop at helping us with our imagination and creativity. Just like you probably can’t imagine a life without your mobile phone anymore, in the future, people might not be able to imagine how people in our time could have had ideas by themselves without the help of their creative assistant.”