Holly Herndon knows that every aspect of today’s technological dystopia really is deeply human at its core. On her 2015 album Platform, the Berlin-based composer fed haunting chorale vocals into custom software tools (like her “net-concrete” Max/MSP patches) to create a dynamic portrait of the human body in flux; with Skype samples, Greek yogurt ads, and eerie ASMR therapy for our Silicon Valley elites, the album proved prophetic in its focus on the infrastructural limits of contemporary humanism, recognizing the dynamic intimacies of YouTube as just another form of care work, and the unsettling elegance of the influencer economy as really just a new face for some of capitalism’s oldest dynamics.
It maybe makes sense then that Herndon would be drawn to current developments in artificial intelligence, and on her new single “Godmother,” the composer has reunited with midwestern footwork futurist Jlin to create a song based completely on machine learning tools. Describing the software as another collaborator in its own right, Herndon says that the song was the result of an AI named Spawn “listening to the artworks of her godmother Jlin” and attempting to translate what it heard into a new approximation of Herndon’s own voice. “This piece of music was generated from silence with no samples, edits, or overdubs, and trained with the guidance of Spawn’s godfather Jules LaPlace,” Herndon’s statement continues.
Despite its conceptual origins, the track itself sounds a lot like human beatboxing, with Spawn immediately imitating the rapid-fire triplets of Black Origami tracks like “Kyanite” and “Nyakinyua Rise.” Stuttering clicks and whizzing white-noise filters collide with pitched vocal approximations as the song cycles through some strange mathematical approximation of a Jlin track. From: https://www.spin.com/2018/12/holly-herndon-jlin-godmother-review/
