“En Tren” is an experiment in creating visual poetry by utilizing a generative adversarial network (GAN) to produce computational impressions of the artist’s anti-capitalist, stream of consciousness writings. This particular poem was written entirety whilst traveling on public transit and/or waiting for the bus and light rail in Phoenix, Tempe and Tolleson, Arizona. The animation was composed from sequences of images created by some out of work e-waste computers that were parsing the lines of the poem into images using deep-daze, a machine learning implementation in python using CLIP and SIREN- using a GAN.

BY  Vav Vavrek (US)


Vav Vavrek is a creator of things and an instigator of events. His works and writing evoke some of the helplessness and anger of the new lost generation. Vav empathizes with sad little unwanted machines and, because of this, he desires to give them the ability to communicate, whether through attempts at language or image. Vav wishes for them a new life as virtual Jackson Pollack(s) and e.e. cummings(s), or really whomever they desire to emulate. Some of these machines have less potential than others, mostly due to their hardware limitations and/or physical damage to the units, but also limitations of the software that is available to load onto them. The oldest and saddest of these units, for Vav, are the most treasured because he wishes to give them a place to live out their later years with a grace and dignity that is uncommon inside our throwaway society.