Curated Music

This music was generated live as a part of Ars Electronica 2020, supported by the Bartlett School of architecture’s Interactive Architecture Lab.

Curated & Optional Music …

A common critique of machine learning processes is that they fall into common typologies of pattern recognition and repetition; identifying and classifying information, but not creatively manipulating the data into something novel.

This ‘system’ addresses this position, and blends analytical ML processes with evolutionary computation and human curation to ‘evolve’ music from a base position into something new and unique.

The system is part of a series of experiments used to ‘evolve’ sounds and music to fit a convolved model of a space; audio analysis determines which musical characteristics of a given model go on to seed the next generation.

This method will show us how musical processes can evolve to suit a given spatial condition, just as they have for thousands of years …