Meet my Generative AI dog, Poolander! Just two days ago, creating these four images of my worrywart friend with a single text prompt would not have been possible.
Midjourney just rolled out the new “Character Reference” feature, which allows you to generate new images that aim to create consistent characters based on an image you provide.
This is a big milestone in AI Art (Yeah, I know, I say “big milestone” almost every week) because Gen AI can now enable you to put the same-looking character in different scenes.
So, whether you are designing marketing materials or creating a comic book for children, you no longer have to settle with with random-looking characters when you use Gen AI.
The tech is still in its infancy, though. Poolander appears largely similar in his various portraits here but other sets of animals I generated didn’t look as consistent.
It’s just a matter of time before Midjourney or another Gen AI app gets it right, and everyone can become their own comic book artist. Or maybe you can create your own digital avatar for different social media posts.
What do you think?
BTW, Poolander is named after the famous male supermodel Zoolander, who got insulted when people said that he had “only one look”. But he did only have “one look”!