| We're in month one of these tools, so I think your predictions are incredibly short sighted. > Like the other commenters have written here, art is a lot more fluid than you give it credit for. And the new workflows will be as well. > Professional illustrators will always have work They're in work because the job takes 10,000 hours to learn and individual pieces take 2-10 hours to create. Not anymore. The dam got busted. > In any case, a big part of the work for illustrators do is in refining the vision of their client and exploring the space of art this way. You tell the illustrator, "I don't like the arm". They come back the next day with something new. You tell FancyIllustrator.AI or whatever the same and you can draw a squiggly and have ten new concepts to refine from. Game over. > This is as much of an issue with the current AIs, you need to ask for something very specific to get a good looking image. Iterating to produce a quality image can take hours. This is month one. Everything is about to be disrupted. Even the big tech companies. This is the biggest singular moment of Innovator's Dilemma in human history. |
What was practitioners in the field of computer generated art able to do in month minus one, how has this changed subtstantively, and do they offer any experience, criticism or understanding of the practice that might still apply? If not, which and why?
I’m as bedazzled by the blinkenlichten of prompts and generated images as the next guy, but I must confess I don’t have the foggiest what those guys were (and probably still are) doing before Dall-E. If the needle really is bending against the post I would take their word for it.