Warning: ChatGPT image generation can be addictive.
I’ve heard from illustrators who have tried various AI software claim it portends the extinction of their craft as we know it. I’m agnostic, so I wanted to see for myself. Would I succumb to its power or use it in moderation? Would I be awed or bored?
My starting point was a group of unspecified animals posing for a class picture.
I spent a total of 10 minutes “making” the images below. As I provided various simple prompts, images began to appear (in about 30 seconds on average). The longest wait was around 68 seconds at the outset when I asked the program to shift from a (default) group of benign cartoon beasts to a more sophisticated “Expressionist” drawing style.
Artificial intelligence, or what I call additional intelligence, is quite awesome to watch. We can use it for good or ill, to make us smarter or dumber. It’s “just another tool,” but more so. It is in our lives already, and has been without our knowing it for a long time. But as yet it is not perfect, and my exercise proves it. (As when I asked for party hats on the rabbits, below.) Its understanding of Cubism and Expressionism differs considerably from mine. You can see for yourself where ChatGPT was deliberately defying me.
It is encouraging that it refused to render in the manner of Norman Rockwell. The more artists that prohibit the scraping of their work off the web, the better. But it still bugs me that it put hats on the lions.
What I want as we thrust into the future is an AI that heeds all my wishes. Is that too much to prompt?
