I asked a computer to learn the sound of human speech.

Voiceover and composer Emma Clarke experiments with machine learning as a compositional tool. Part of the RNCM's PRiSM 'Unsupervised' project.