Throughout human history there are examples where nature is the essential inspiration for art. It has been the same for technology. Today, systems are being developed to help us understand nature better, and make use of its faultless ideas for our interests. In this context, superorganisms (species which exhibit social behavior such as ants or bees) are important for the development of computer systems. Social species provide a model of distributed intelligence used in cybernetics. This distribution of independent units combine to constitute a greater entity can also be a metaphor for an artwork; in music, for example, a symphony is composed by tiny elements: melodies, timbres and sections that interact to constitute a greater unit. These natural systems may also be imitated to become the creators of the music itself. In a recent interview with Mijael Gutierrez, Chair of the Digital Music Engineering department at the ITESM, in Mexico City, he spoke about his research on artificial intelligence using distributive systems for music creation, and how it was an opportunity for him to integrate his backgrounds in science, music, and engineering.
Mr. Gutierrez argues that most works using artificial intelligence for music have been developed using centralized systems, such as neural networks and cell automata. His pioneering work in using a non-centralized system for automatic music creation, which, according to Marvin Minsky (whose theories Mr. Gutierrez bases his work), is the most successful for programming algorithms for creative purposes. Distributive systems also happen to be more economic in terms of computer resources.
To construct a system that composes music in a similar way that humans do, human creativity must be studied. The process by which humans learn, interpret and appreciate music is extremely complex; it is again a collaboration of a number of organs and processes that serve for a greater task. This actually turned to be the most difficult part of Mr. Gutierrez project, to create an algorithm that analyses music and imitates styles, basically he is looking for a music student. A melody, for example, has an enormous impact in people we can immediately conceive of it as a separate element of a piece and memorize it. We can listen to a melody and then keep it in our minds and whistle or sing it weeks after we heard it. As natural as this seems it is extremely complex to analyze the elements that constitute a melody, not only from the cognitive perspective, but also from a music theory perspective. Many attributes are to be considered, rhythm, pitch, time relationships, harmonic relationships and many more. Then, we have to consider how we react and perceive all these attributes from a cognitive perspective. To create an automata that creates only melodies is already an extremely complex task. Of course, all the emotions involved in the composition process are subject of great debate of whether an artificial system could be capable of composing music of human quality. For Mijael Gutierrez all these subjects of cognition and emotions are also variables, although he says that this does not have to denigrate nor demystify the work of a composer, but only help to understand human’s creative process and serve as a tool for artists and entertainment. He also states that it will be long before we can imitate human’s composition process, and that artificial intelligence is as intelligent as we teach it to be.
All this research, as well as Mr. Gutierrez’ background has an impact in future generations of artists and scientists for whom the barrier that differentiates both of these disciplines is being torn apart. For him, we differentiate between these two only because we are taught to do so, but fortunately the world is opening more and more to multidisciplinary developments and additionally institutions are being founded based on these principles. One of his objectives is to form students that do not have to choose between becoming scientists nor artists, but they may be both at the same time, “it is all part of the same”, he says, “but we have to stimulate them”.
The whole idea of imitating human creativity in computers is a subject of great debate and controversy between art-philosophers and science-philosophers, and between artists and scientists. This goes far away from the scope of this article, but the sole conception of social species as a model for music is of enormous aesthetic richness and is a source for philosophical, artistic and scientific, creation and is clearly worthy of study and experimentation. The abilities of non-centralized systems have been proven by nature already, and we have this concept to our disposal to create and develop new ideas and paradigms.
Latest posts by Marte Roel (see all)
- Tania Candiani: Five Variations of Phonic Circumstances and a Pause - April 13, 2013
- Contemporary Creators Archive, Mexico City - April 21, 2011
- Superorganisms as a Model for Music - February 7, 2011