The poem in the video below was inspired by Langston Hughes’ Island. It was developed by running it thorugh a series of Google Translate operations and also an N+7 operation. Then editing. I think this is an example computer-aided composition without using AI (though I suspect—OK, I know—there is AI in Google Translate.) There is a difference between LLM-using generative AI and computer-assisted work. For me, the process “makes things strange” and offers up some textual materials to explore, all of which is infused with the aura of the original. How I respond to the materials and the aura is part of my poetic process. For me, these aura-filled texts often confound and inspire and my work as a poet is to someone work with the material to arrive at some genuine “poesis” —the sense that the text is more than the sum of its parts. I don’t know that it is that much different than writing after being inspired by a poet or poetic tradition, other than generating some related material to respond to. I should say that the poem that I settled on it very traditional and doesn’t confront the norms of conventional poetry except in its mode of composition. I’ve been using computer-aided operations since the 80s, though definitely not exclusively. I don’t see this as much different than rolling up one’s sleeves and addressing the affordances of a sonnet or villanelle or Frank O’Haraian “It’s 12:15 and this and then that” convention.
What a beauty of a poem by Hughes.
Another thing that interests me in this video is that the beginning has the text and then most of it has the solo saxophone responding to the text.