Computer Character Book

Computer Character Book was an art and research project into mixed-class image generation using AI made in 2016. A fundamental aspect of human creativity involves mixing symbols and meanings in new ways, to generate fantasies, metaphors and imagined realities. Most existing artificial intelligence programs facilitate the association of a single word to a single image in a 1:1 relationship but I was interested in constructing an AI that is able to sporadically merge concepts it has only previously learned separately.

My research into mixed class generation using a new CNN-RNN autoencoder is discussed in a more detailed post here. The model is inspired by the way Chinese characters are encoded both spatially and in stroke order. In order to create an AI capable of dreaming, I made an AI capable of encoding drawings in both space and time.

I present the final work as a 10-page booklet of fantastical imagery, meant to evoke the character manuscripts of Chinese handwriting where complex language was developed through imagined images. Some pages were drawn using a pen-plotter blot and other pages are printed on rice paper, withe book being bound using dyed cardboard.