Forms and Models of Training
multi-media installation,2020-2021Forms and Models of Training uses AI StyleGAN, an artificial intelligence technology, to analyze thousands of released photographs of Olympic medalists from the Internet Archive, which have a very high degree of anti-duplication in their screen composition, and ultimately generates an infinite number of non-existent medalists.
The generated fictional medalists maintain a similar screen composition, posture and facial expression to the original images, and can be said to be models that embody the intangible concept of "victory.
This process of repeating the same thing in an unhurried manner can be called "training" for machine learning, just as athletes repeat their daily practice to win medals.
Compared to the cost of producing a real medalist, a medalist generated by AI technology could be more efficient, more reliable, and more useful. Isn't this the kind of rule that should be paramount in today's production activities?
exhibition history:
‘Triangular Relationships’ 3331 Art Chiyoda, Tokyo,2020
‘PANACEA’, Calm&Punk Gallery, Tokyo,2021
‘Guerrillas in Flatland―Emerging Curator Project’ Shanghai Modern Art Museum, Shangai,2021
‘Man-made Landscape ― Hotel Asia Project ’Chengdu Time Art Museum, Chengdu,2022