Can animal and artificial intelligence become artists?
Can animal and artificial intelligence become artists?
To whom can we assign the epithet “creator”? Are animals creative? And how is creativity cultivated by artificial intelligence (AI)? What do we actually mean when we talk about creativity?
The current MO Museum exhibition Animal – Human – Robot is complemented by various events aimed at analysing the link or relationship between humans, animals, other life forms and technologies. One of the lectures held at the museum was delivered by Skaidra Trilupaitytė who attempted to target the question of creativity of various life forms and technologies. The following article highlights some key ideas of her lecture “Person in the Process of Creation: Between Animal and Algorithm?” as well as provides some insights regarding some practices and examples of non-human artwork.
Animals that paint pictures
During the opening weekend of MO Museum’s exhibition Animal – Human – Robot, I attended a lecture on human and animal psychology by Rūta Sargautytė. In this lecture, she said that pigs are some of the most conscious animals – they can recognise their reflection in the mirror. While studying this story, I discovered that pigs are very curious – they explore the environment by grabbing, tasting or pulling. One lady has taught her pet how to paint pictures. The result of this were artworks similar to those of abstract expressionism, and the pig itself was given the name Pigcasso.
The mating rituals of animals also evince their sense of aesthetics. Not only many species of birds build their nests comfortably but they carefully decorate them too. Some more primitive creatures (like fish or bugs) use a variety of matchmaking tricks on females. Therefore, it is not surprising that the “aesthetic” ingenuity of some animal species still catches scientists by surprise. Animal songs or dances usually serve a pragmatic purpose of extending their kind. But, as I said, there is also spontaneous curiosity or some practically “aimless” behaviour in the animal world. Painting can have a therapeutic effect on animals, just like on humans – creative activity improved the well-being of the pig mentioned earlier.
There are also other reasons highlighting the similarities between an artist pig and an artist human being. Pigcasso did “artistic work” for food, and so do many artists we know. Understanding that some of the most widely recognised geniuses did their paintings for a dinner or place to stay overnight does not diminish the artworks’ value for us, the spectators. And if creativity is broadly described as being able to solve a problem in an original way, one has to admit that in unexpected situations, both humans and animals tend to mimic or deliberately imitate and sometimes even find unconventional solutions.
Art industry is diving into AI
Since new technologies have been validated as a field of expression in the art world for already a long time, I cannot escape talking about them when it comes to creativity. Nowadays, a number of programmes generate different styles of music or images, and there exist algorithms trained to perfectly mimic people, including well-known artists. It is important to emphasise that computers are now much more powerful than twenty (or more) years ago, and today’s achievements in AI products are marked by the unprecedented access to databases with hundreds of thousands or millions (in some cases billions) of diverse data. And in recent years, machine learning and deep learning of algorithms have opened up unprecedented creative spaces.
Readers of Yuval N. Harari’s bestseller Homo Deus probably remember the story of David Cope, a controversial US musicology professor at the University of California, Santa Cruz, who created computer programmes that produced concerts, symphonies or chants. The speculations about the authorship of music did not end for the benefit of the people – the connoisseurs found themselves greatly annoyed as, based on their “competence” (unfortunately by mistake), they attributed some pieces that were actually created by the programme to famous composers. However, the fiasco of human expertise should probably no longer come as a surprise. Even the most receptive human mind in the endless stream of information is no longer capable of “sorting out” its own taste receptors or recognising its individual needs. That is why recommendation engines are being developed that classify the output of Netflix, Spotify and other platforms according to individual interests. Smart devices subtly “improve” live sound, adjust and even “create” beautiful photos, recognise images, motion or identify additional information based on visual noise. Despite the problems encountered in these processes (biased information selection, gross discrimination errors), companies are constantly improving their products.
The high art industry has also dived into the field of AI. In 2018, one of the world’s oldest art auctions Sotheby’s bought the AI starter Thread Genius, the developers of which have previously worked for Spotify. It is a suggestive algorithm that identifies art objects and can offer artwork to individual needs according to various criteria. What is more, with the help of Sotheby’s Database (with a history of over 50,000 deals), the Thread Genius programme “knows” the regularities of price fluctuations and provides a fairly accurate estimate of the future financial value of the artwork and the potential change in the investment value.
Today, by attending events such as the Venice Biennale, you will find more than one object that, according to the explanatory notes, was created not only (or not thus much) by artists but by a programme collaborating with a person.
And who owns copyright?
Non-human authorship today is well illustrated by a case from the animal world – the story of the selfie taken by macaque monkey Naruto. An anonymous big monkey in the Indonesian jungle became famous and received a name after one day it found a camera left by the nature photographer David Slater and made a selfie that the photographer published in the media in 2011. Slater began selling the photo taken by the monkey and retained the image’s copyright. With regard to this, PETA, the organisation protecting the animals’ rights, sued the photographer for unlawful misappropriation of the picture’s copyright. Various organisations (even Wikipedia) have been involved in the controversy, and the intensification of the animal rights discourse has resulted in autumn in 2017 – the process ended in Naruto’s favour. Still, in 2018, the court ruled that the animal cannot own the copyright. However, we certainly cannot rule out the possibility that future courts will look at animal authorship in a different way.
In autumn in 2018, Christie’s auction announced that it would be selling works produced by AI. The main piece was Edmond de Belamy’s portrait blurrily depicting a “French gentleman” dressed in black – a representative of the fictional Belamy family. At the auction on 25 October 2018, the painting’s sale price exceeded its initial expectations by more than forty times (with the final price standing at USD 432,500). This sensation led to speculation regarding the reasons for its success. The fact that the portrait was displayed alongside the work of Chuck Close or Jeff Koons caused confusion not only in the world of art but also in the world of programmers. The creative process took months whilst the generator was “playing” with the discriminator against the generator in a system that was loaded with 15,000 portraits dating from the 14th to 20th centuries, with the final result being “signed” by an algorithm. Those who are at least a bit AI-minded soon found the idea that, in technical terms, essentially was neither very original nor very sophisticated. Meanwhile, the final price was driven by high expectations from the public, a lack of programming awareness and a public relations message claiming the work was created by AI.
Following the success of Christie’s, their competitor Sotheby’s has also taken on a similar ambition. In early March 2019, it introduced Mario Klingemann’s work “Memories of Passers-by” – an installation consisting of a wooden console with an AI brain and two screens. They were projecting an endless stream of unpredictable associations between female and male faces that were created and “released” in real-time. This, according to the Sotheby’s spokesperson, gave viewers a unique opportunity to see directly “how the brain of AI was thinking”. Unfortunately, this piece did not promote the rush that marked the sale of Belamy’s portrait – Klingemann’s installation was sold for GBP 40,000.
The question arises whether an algorithm can be described as “creative”? Cannot we apply this concept to it just because it has no subjective consciousness? Today, not only the idea of imitation but even the genius who rebelled against the system and broke the rules can cause suspicion. By studying the chemical processes that take place in the brain and subsequently manifest themselves in different actions and reactions we come to the conclusion that both spontaneous human feelings and deep beliefs are driven by both natural and artificial systems. Invasive materials, new formulas, algorithm suggestions, viruses, linguistic games, etc. are constantly changing the boundaries of both humans and AI.
Creative human and longing for imperfection
With regard to the progress of the intellect, it is pointless to continue the confrontation between the competences of humans and AI. It is likely that Homo sapiens will increasingly work with AI in the fields where expert judgment is needed, and it is highly likely that our intentions may be increasingly intertwined with the capabilities of non-biological systems. Yet, in the creative space, there exist experiences that are unique to humans. In this case, it is not about the pursuit of perfection but, on the contrary, about breaking the rules, the inexorable desire to remain imperfect and the imperative of destroying mathematically modelled modes of rationality. We would be wrong to assume that artificial modes can be totally predictable.
In the systems, millions of interacting algorithms sometimes “mix” into toxic loops that are inscrutable for the human mind, resulting in unpredictable and, sometimes, disastrous results. Errors can be extremely expensive in the fields of medicine or transportation. However, in the space of diverse human experience, creative destruction, as well as the enigmatic concept of true art, is a socially constructed phenomenon. And if we agree that the essence of creativity and novelty is the rejection of traditions and conventions that leads to a certain shift in paradigms, going against mainstream or even the new concept of a human, then for our kind creativity will remain a prerogative.