CHO You Kyung “AI Music from Aesthetic Viewpoint”

My research life started in 2008, when I encountered a study on “Aesthetics” in Japan.

It is not easy to define “aesthetics” in short. If I would like to explain straight forward how human beings acknowledge ‘beautifulness’, and ‘sensibility’ at that moment. So, every deed done by human beings can be an object, and its scope is infinite. For such reasons, the center of discussion about ‘aesthetics’ would sometimes be ideology or art, depending on the time or country. As the definition of art is being diversified, as mentioned above, it is possible to make research like this essay. As an example, let’s study AI music from an aesthetic viewpoint.

 

If you put a lot of music into an AI system, it would make similar and new music by analyzing original data. For example, “AI Composer Emily Howell” composed by David Cope, can produce a lot of music in a short time based on the old composing style of Beethoven or Mahler.

In the field of popular music, a research team led by Ahn Chang-Wook under the Gwangju Institute of Science and Technology in Korea developed “AI Composer EvoM”, and various popular songs including K-pop. Against such AI music, regardless of academic or non-academic field, public opinion has shown its interest in such scientific and technological aspects of AI as its historic, scientific potential and commercial value. Moreover, such controversies have recently extended to the field of ‘post-humanism’ like social and ethical issues, and aesthetic and philosophical fields by environmental philosophers.

 

We can re-consider, from the relationship between AI music and human beings, “creativity” which is one of the main concepts of “aesthetic”. In the evaluation of AI music in classical music (western artistic music), it is criticized as poor tasting songs that imitate songs by great classical composers simply. When I listen to AI music, I think its quality is much inferior to ‘music by human beings’. However, such criticism may be caused by the creativity of ‘human music’, which has been controversial since the latter half of the18th century. At that time, the concept of ‘genius’ or ‘originality’ rose, and the composition using existing music, which has continued since the Middle Ages, was criticized.

 

Namely, borrowing behavior itself means no originality. For example, the citation technique of Gustav Mahler, composer of the post-romanticism era, was related to ‘Jewishness’ (negative meaning), and its originality was suspected. Such a tendency lasted until the modernism of the 1950s and 1960s. For composers and critics who asked for novelty, the use of adjusted music means ‘return to the past’, and quoting the behavior of existing music means ‘dirty music’.

 

Deliberation for ‘creativity’ in AI music may become the start of this issue. Against various aesthetic questions for AI music, like 1) What is creativity? 2) What is the function of a composer? 3) What is work? 4) How do listeners react? The answers would be the following: When you input information into AI, it is ‘human composers’ that select information, and, on the other hand, it is AI that creates new music. As David Cope mentioned above, even if ‘human composers’ collect, select, and input information, AI will produce unexpected results.

 

From the theory mentioned above, I can say that AI has ‘originality’. According to David Cope, it does not depend on human inspiration alone but arises from another factor like a machine. And ‘originality’ does not come from anything but the context that comes around ‘originality’. ‘Originality’ comes from the composition of works that other people have done. Furthermore, it must rely on the judgement, for or against, of other people who accept or reject the things that are aesthetic. We can find such insistence of Cope in the process of creation, not in the result of works of ‘originality’. Such discussion about the ‘originality’ of AI marks an important beginning for understanding the ‘originality’ of human beings. It gives us the big suggestion that we create something new from nothing, and we can review our attitude toward appreciation, which is bound by modern mythology.

 

 

SGRA Kawaraban 739 in Japanese (Original)

 

 

CHO You Kyung /2021 Raccoon, Postgraduate student at the University of Tokyo

 

 

Translated by Kazuo Kawamura

English checked by Sabina Koirala