In his first solo exhibition in ten years, Making Sheep, Hirofumi Katayama will present new works using the technique of adversarial generative networks (GAN), which is image generation by artificial intelligence. All of the works on display are images generated by using a large number of Katayama's photographs as training data. Going back to 2013, Katayama presented the 'Vectorscapes' series, in which he created 'fake photographs' by importing photographs into a computer and converting the images into numerical values. The new works in this series, although using different techniques, repeat the theme that Katayama has been pursuing, namely, the creation of fake photographs.
Katayama says that the creation of images using artificial intelligence is interesting in that it is neither a simple mechanical operation nor an artificial act and that an important element of his work has been his interest in the human obsession with the concept of the 'real thing.' Can the 'real' be confirmed only by comparing it with a 'fake?' Can a photograph be called a 'fake of reality?'
In this age when the reality of photography is rapidly disappearing due to artificial intelligence, Katayama's new works lure the viewer into a complex and profound labyrinth concerning the existence of the 'real.'
Press release courtesy Taro Nasu.
Piramide Bldg. 4F
6-6-9 Roppongi
Minato-Ku
Tokyo, 106-0032
Japan
www.taronasugallery.com
+81 035 786 6900
+81 035 786 6902 (Fax)
Tues - Sat, 11am - 6pm
Closed on Sunday, Monday and Public Holiday