In doing my research, I read many books about technology, interface design, and theory. I typed many pages of exact quotes from my reading. In addition to these, I also collected text from science fiction novels and short stories, Catholic prayers, and my own writing. Though the texts seem disparate, I intuitively knew they fit together. Some of this content I consumed deliberately and some not, but they are all significant influences and are inseparable from who I am as an artist.
Using a process analagous to how I make my sound and video compositions, I wrote code to algorithmically generate new text from these carefully curated sources. The computer program breaks down the input text into a soup of sentence fragments, and it then combines the fragments together to produce pages of new sentences. Much like interpreting the inkblots of a Rorschach test, I read through these output pages to pick out the sentences that stood out as meaningful. After selecting the sentences, I limited my editing only to correcting grammar and changing punctuation. I did not change the words. Finally, once I had many of these sentences, I rearranged them to produce new writings in the only way they seemed to make sense.
Traditionally, a writer does research and connects existing words in new ways to build meaning, and the above method is not so different. I am simply using the computer as a tool to aid me in finding these connections.
I do not believe that these passages are accidental or nonsense. Like an archeologist uncovering pieces from some kind of shattered object, I am unsure of what I have discovered until I put all the pieces together, and these pieces can only fit together in one way. The content of these generated passages was already there. I only had to put the pieces together to understand the whole.