patten announces  first ever LP made from text-to-audio AI

“An emotion-drenched, trippy joyride through contemporary music, and a radical glimpse into the near future and beyond.”

Made using ‘Riffusion’ “a neural network, that generates music using images of sound rather than audio” and utilising what it’s learnt from its vast dataset of recorded music history.” patten used hours of lo-fi recordings from these sessions for tiny sections with “countless hours of chopping, sequencing, layering, EQ, repitching and fx”

Across its 21 beat-tape-meets-Musique-concrète style tracks, genres from pop to techno, R&B to opera, hip-hop and grime to ambient, are distilled into half-remembered fragmented forms.