While looking for images to tell the story of each Strange Read episode, I’ve discovered some beautiful and haunting paintings from the past. However, by far the most chilling have been some images from the future. Yes, the future.
I stumbled upon AI-created artwork while reading a post by Nekro. This artist produces new pieces by asking a computer to generate an image and then finishing the composition manually. It couldn’t be available without thousands of dollars of software, I thought to myself. Despite my reservations, I did a casual search. What followed left me stunned.
One YouTube video and a link later, I found myself at the VQGAN+CLIP.ipynb site. Type a description of what you want, and your computer goes to work. Many of the AI art examples are of colourful future-scapes. However, following my instructions, my PC soon belched out images from the far reaches of hell.
At 12:30am, I sat across the room, watching my monitor at a distance. As iterations scrolled up the screen, a sense of dread – like the beginning of a horror movie – consumed me. In a podcast several years ago, we laughed at ourselves for describing how VR headsets were mind-blowing. It feels horrible to talk of new technologies, knowing the future might listen to our foolish words.
Creating AI images feels both fantastic and horrible at the same time. The images are intriguing. They’re immediately useful, and no copyright is involved. However, one worries about the human role in creating any art. Words, music, and video are all possible AI creations. What happens when they’re better than us? Will they eventually appreciate and interpret the art instead of us?
I’ve never been one for paranoia about the machines taking over, but last night, I think I got my first sense of what it all means.