Sam Scorsby's 'Impermanence Series A' generates a new, unrepeatable audio album every day by stitching together fragments from shortwave radio, FM broadcasts, web streams, live microphones, and Internet Archive recordings using Python scripts he has published openly.
The project is a deliberate experiment in human-machine collaboration: the software handles acquisition and assembly, Scorsby listens daily and selects his favorite tracks, and the full run will end in a human-curated 'Best Of' collection. His accompanying essay argues that computers process raw data well but lack the contextual judgment a human brings, and that argument is worth reading on its own terms, separate from the audio project itself.
Whether the output qualifies as music is left open. What makes the full write-up worth your time is the technical breakdown of how the pipeline works and Scorsby's unusually grounded position in a debate that rarely produces grounded positions.
[READ ORIGINAL →]