[ISEA94] Paper: Carsten Bredanger – Expert Systems: Building a Bridge to Electronic Sonic Art

Abstract

Despite, or perhaps because of, a healthy skepticism, Artificial Intelligence (Al) has been making quiet progress in electronic arts. Artificial Intelligence has inspired traditional fields of electronic arts as well as it has developed new horizons composition for many artists working in electronic environments. Building on the success and shortcomings of previous experiences with computers in arts, the attempt to extend the paradigm of artificial intelligence systemsto the domain of electronic sonic arts is made now. Musicians are increasingly using intelligent machines to deal with tasks for which they are better equipped than humans. Computers are increasingly being used to address the brain-numbing complexity of modern electronic music products and processes, thereby allowing people to concentrate on their music and ideas. Expert systems, for example, help people by searching a book of rules to decide what to do in a particular situation; as machines do not forget, these systems can manage rules more consistently than people. Some musicians are using neural nets, which can recognize complex patterns, to apply precedents that are difficult to express in numbers or words.

The real challenge facing technology is to recognize the uniqueness of machine intelligence and learn to work with it. Given enough memory, a computer can remember everything that ever happened to it or to anyone else. Furthermore, when faced with a logical problem or a theoretical model of how compositions or sounds should be, computers can deduce more results more quicklythan humans. Their complementary strengths should allow man and
computer to work together and do things that neither can do separately.
Artificial Intelligence has already shown ability to help musicians work and gain access to all the computer power in order to deal with increasingly broad soundprojects. The practical uses of Artificial Intelligence, expert systems, are presented and explained, e.g. some of the different ways in which machines can be used for reasoning or processing vision and language data in computer systems to improve the interaction between humans and computers. Central
to these computers is the element of pattern recognition, an element of human intelligence that can be thought as a kind of a detective that tirelessly collects and recognizes visual and auditory patterns.
The ways in which musicians and computational-sound-environment users work with AI are discussed. In tables and sidebars, a glossary is presented; sound applications that benefit from
Artificial Intelligence are listed; silison neurons are described; and the costs and computational densities of the human brain and of expert systems are compared. Artificial Intelligence techniques require a lot of processing power and computer memory, but the driving force
behind the use of more intelligent computers is more intelligent artists and computer users. It should be remembered that Artificial Intelligence, like any useful technology, is no panacea, as
there are occasions when the technology just does not work or is badly managed.

  • Carsten Bredanger was born in 1970, studied philosophie, germanistics, computational linguistics and artificial intelligence at University of Cologne and Duisburg (both in Germany), Leuven (Belgium), musicology and composition in Essen and Duesseldorf
    (both in Germany). The list of works contains instrumental music, electro-acoustic music, multimedia and video, which are noticed internationally. Owner of the “Image Acoustique Studios”, Duisburg.