deep convolutional general adversarial networkcuneiform-tablets; code; laser-cut acrylic;

installation; 6 x 8 inch;


c o n c e p t

take the human element out of a human process; create new hyper-artificial Sumerian cuneiform writing with neural networks;


k e y  p r i n c i p l e s

neural networks, machine learning, human code, algorithmic writing, generative art, post-human communication. 


p r o c e s s

Cuneiform is regarded as one of the earliest systems of human writing & encoding of information. It was common during the Sumerian period in the fourth millennium BC. A blunt reed was used to stamp words and letters into small pieces of clay, the tablets were burned afterwards. I trained a DCGAN-neural network on about 20 000 transcriptions / drawings of cuneiform tablets. It generated new hyper-digital versions of this ancient form of encoding. I took the generated samples and visualized them in different ways: For the image above a generated sample of the network was digitally altered in Adobe Illustrator, for the video below I used processing (code is based on a processing example, image extrusion) for a number of generated images. As a last iteration I laser-etched the cuneiform into neon-colored acrylic - a modern version of the physical form of cuneiform.