It’s interesting that this is a style that’s relatively easy to get AI to output, yet AI slop is instantly recognizable while this actually looks like good art.
Draw me an illustration of the following. A lady with pink, long flowy hair holding up a blank sign. She is wearing an off-white silky gown. The background are some pastel coloured pink lilies, and sky in the colour of pastel blue, yellow, and green.
Sign should be above her head. The lady has Caucasian features.
She should have a face that looks angry and determined to right a wrong. Also, use some bold outlines.
Sure, it doesn’t exactly have character like the one in the OP.
But I also put almost no effort in prompting (less than a minute). And Bing (Dall-E) isn’t exactly the pinnacle of image generation AI. And the whole tech will only get better by time.
The problem is that it doesn’t even have to be “very good”; the majority of people are entirely happy with “good enough”. It’s not hard to imagine how the whole thing will threaten and disrupt so many industries.
Compounding on the “good enough,” i feel like when someone reaches out for a commissioned piece, they have an idea in their mind what they want, but are terrible at explaining it. So what the artist creates might not be what they wanted, with AI, they can make a prompt, see that their explanation sucked, and change their explanation in faster cycle time
It’s interesting that this is a style that’s relatively easy to get AI to output, yet AI slop is instantly recognizable while this actually looks like good art.
I had some fun with Bing AI…
Sure, it doesn’t exactly have character like the one in the OP.
But I also put almost no effort in prompting (less than a minute). And Bing (Dall-E) isn’t exactly the pinnacle of image generation AI. And the whole tech will only get better by time.
The problem is that it doesn’t even have to be “very good”; the majority of people are entirely happy with “good enough”. It’s not hard to imagine how the whole thing will threaten and disrupt so many industries.
Compounding on the “good enough,” i feel like when someone reaches out for a commissioned piece, they have an idea in their mind what they want, but are terrible at explaining it. So what the artist creates might not be what they wanted, with AI, they can make a prompt, see that their explanation sucked, and change their explanation in faster cycle time
just look for “Alphonse MUCHA” online, you’ll get all the answers you’re looking for
For now. I worry it will get to a point where we won’t be able to tell.