THE BEDSIT ROYAL: Exiled princes and other minor royalty of dubious provenance have a peculiar status in American business, literature, pop-culture and art. Royal titles are valued and keyed to ethnic identity (witness the plague of Cherokee princesses and Irish kings, for example) but they are also despised and justly associated with fraud. Every few years a new royal fraud pops up (most recently Anthony Gignac, a Colombian orphan who spent years pretending to be Prince Khalid of the Saudi Royal family to demand credit and special treatment). To explore and inhabit this weird phenomenon and interrogate an extreme form of white privilege, I used a machine learning model to generate images of a lost kingdom and coined a cryptocurrency for the lost realm (lost empires such Rhodesia or Northern Ireland have a similar cultural weight to royal titles.) Embodied in this project is the idea that artificial intelligence is anything but artificial, rather it’s similar to divinity in that it’s a palimpsest, human activity (in this case 500 photographs of royal interiors) that have been processed by a pair of adversarial algorithms, a generator and a destroyer (in this case 16,000 times) to create a series of new images such as the one above. I selected the images then improvised vignettes based on the images in a faux 'sales' setting. Viewers who took the currency could trade it for a real ink drawing based on the images, creating an exchange and bringing the currency and by proxy the lost kingdom to life—James McGirk