“A beautiful, lush jungle with high, ancient temple structures. There’s a clearing in the jungle in front of the temple.” After a few seconds, structures and textures start to appear from what looks like a noise filter in Photoshop gone rogue.
The temple in the jungle is right in front of me. Generated by a computer, based on the prompt I wrote, and materialised from training sets of millions of photos and pieces of existing artwork.
The world of AI-generated images is still relatively new to me. I’ve seen examples of it over the last few years but never really invested any time diving deeper into the topic until recently. Nvidia Research tools allow users to paint suggestive, simple brushstrokes that the AI turns into landscapes. Artificial human faces that look so real, apps on my phone that can generate images from nothing but a few lines of text, and tools that create realistic portraits of cartoon characters and vice versa.
There’s no doubt that AI will change how we make images and tell stories. In the best case, AI will augment human creativity and enable us to tell stories in ways we could never imagine. In the worst case, clients will turn to machines to get the job done faster and cheaper.
The ethical implications of thisor even more significant than the advent of the camera or digital art.