This small set of ‘extractions’ is the product of models created from fragments of images from ’90s advertisements, newsreels, and old CCTV footage. The models were combined with key images from their dataset and prompts in an attempt to ’distill’ them down to their core. The images were then sequenced and processed for video.