Floor Location : M 098 D
Since the beginning of human existence, music has been cherished by humans. The creativity of the human brain has allowed humans to write music that displays emotion, tells stories, and brings people together. In the rising age of technology, computers are being used for many tasks previously done by humans, and are doing it in less time and in better quality. This is an experiment in machine learning to find out if artificially intelligent learning machines trump creativity when it comes to composing a piece of music. Computers learn by randomly doing something and then being told whether it is good or bad. The computer will then design rules based on the feedback. The human was given 10 minutes to come up with 6 melodies each with 6 notes. Then for the machine, 6 melodies each with 6 notes were picked using the random note picker. These melodies were judged blindly by a music theory expert. Based on the judging score the random note composed melodies were given rules they must follow when composing. With the new rules, the machine composed again and received a new score. This process was repeated 10 times. The first few rounds of the machine composed melodies showed an increase in score, but the later rounds found a decrease due to lack of originality. In all cases, the human-created melodies scored higher than the machined created one's. It was concluded that machines can improve their composing through implantation of rules but it is not enough to override human creativity.