Will music improvisation make AI even more intelligent?

Aug 11, 2015 15:27 GMT  ·  By

The future DARPA robots, which will probably look like Terminators with a soft side, are taught to play Jazz in order to test spontaneous, real-time adjustments in changing patterns of behavior.

And the truth is that it doesn't really matter if the actual reason behind this research of robots and jazz is really "spontaneous adjustments in changing behavior" or not, as it is after all a very interesting project that wants to close the gap between the robot brain and the human mind in one way or another.

The way the researchers are planning to "teach" robots how to play music is by compiling a database of thousands of transcribed musical performances by the best jazz musicians of all time. Then, using all sorts of algorithms, the AI will understand how to detect successful ways of improvising new melodies.

The new research is indeed unique, as it wants to bring robots closer to the ability of creating music by improvising on known musical tracks made by real musicians. While this improvisation will probably be made to follow a certain musical pattern, with all sorts of other influences mixed together, it's the AI who decides how to mix the songs in order to create something close to jazz.

According to TechInsider, the project is led by Kelland Thomas, a jazz musician and computer scientist that aims to make computers as creative as humans in order to increase the collaboration between man and machine. Why did he choose Jazz? Because "in my mind, jazz and improvisation in music represent a pinnacle of human intellectual and mental achievement," Thomas says.

Not being directly linked to this Jazz project, here's the "Compressorhead" robot band rockin' it sci-fi style.