Faces of ImageNet is a warning about the dangers of artificial intelligence with reference to the ImageNet databases, a collection of more than 14 million images divided into around 20,000 categories and used in artificial intelligence learning and computer vision algorithms. The interactive installation captures the face of the participants placed before a camera and projects the tags that the ImageNet learning algorithms use to classify them onto the façade. The result is not always flattering and can reveal the biases, prejudices, racism and pseudoscientific premises that run through many artificial intelligence systems.

With the collaboration of Fundació Foto Colectania

Trevor Paglen

The North American creator, writer, photographer and geographer, Trevor Paglen, reflects in his works on the gathering of data, the invisible power of governments and the surveillance to which the population is submitted. Born in Camp Springs, Maryland in 1974, he obtained his PhD in Geography from the University of Berkeley and studied art at the Art Institute of Chicago. Paglen has exhibited work in cities around the world and his latest pieces address the formation of artificial intelligence through the images and criteria with which they are defined.

During his time as a writer, he wrote Torture Taxi, in which he describes the CIA’s kidnap and torture programme, co-written by the journalist A. C. Thompson; I Could Tell You But Then You Would Have to be Destroyed by Me, in which he addresses the system of secret military projects not recognised by the United States Government, and Blank Spots on the Map: The Dark Geography of the Pentagon’s Secret World, which looks into the opaqueness of the United States Department of Defense.