AI? Oh My!

In April and May of 2023 I was a beta tester for Firefly, Adobe’s AI image generation program. I generated hundreds, maybe even thousands of images. Not surprisingly, I also generated some thoughts about the process, using OldSkool Intelligence® (the kind inside our heads 😁).

AI generation of whole images is a new artistic medium. It is analogous to the introduction of photography in 1826. Despite the fears of painters and other artists, it did not replace them and their work. It was, undeniably, a disruptive technology – just like AI today. It changed the way many artists thought and worked, and undermined some revenue streams (e.g. portraiture), but painting, drawing, printmaking, etc., did not disappear. Indeed, these mediums went on to scale new heights. I no longer have doubts that photography will do the same.

Having worked in both mediums, I have seen that the underlying skill sets and artistic sensibilities differ. High quality “artistic” AI image creation starts and ends with words. Language skills are paramount for both conceptualizing and realizing your image. Creating good photographic art begins in visual understanding and perception of the world. Actualization is based on a deep understanding of the craft of photography.

So, here are some of the more interesting images from testing Firefly. I want to point out that getting the initial image usually required many tries and much refining. And none as presented here are straight out of text. All have been extensively altered and edited. That included cropping, reframing, adding, subtracting, or moving elements, altering colours, adding effects, and on and on. Most of that work was done in Lightroom, Photoshop, Boris FX Optics, and Topaz Studio, rather than in the original AI. I doubt any of these took less than an hour in post.