In AImbient Niklas Dahlqvist (DDM, dirty oil) utilizes a variety of machine learning
algorithms to investigate questions about authorship and simulated collectives.
The music is mainly composed with neural sound synthesis algorithms like sampleRNN and WaveNet.
Using a corpus of contemporary electro-acoustic works as input, and by
intentionally sabotaging the training process, Dahlqvist crafts artificial
intelligence-powered pseudo-lofi walls of sound paint. Embracing the skewed
nature of these badly trained algorithms, the music is on full blast at the same
time as it is fully shut down.
By exploring and misusing neural networks, Dahlqvist raises questions about authorship and the role of the composer in the age of machine learning. Instead of being a means for reconstructing, the warped algorithms become a medium for simulated music, that relates to the pieces in the dataset, whilst not striving to imitate it.
>>> To album
algorithms to investigate questions about authorship and simulated collectives.
The music is mainly composed with neural sound synthesis algorithms like sampleRNN and WaveNet.
Using a corpus of contemporary electro-acoustic works as input, and by
intentionally sabotaging the training process, Dahlqvist crafts artificial
intelligence-powered pseudo-lofi walls of sound paint. Embracing the skewed
nature of these badly trained algorithms, the music is on full blast at the same
time as it is fully shut down.
By exploring and misusing neural networks, Dahlqvist raises questions about authorship and the role of the composer in the age of machine learning. Instead of being a means for reconstructing, the warped algorithms become a medium for simulated music, that relates to the pieces in the dataset, whilst not striving to imitate it.
>>> To album
