I have not tried this service, but one of the examples takes a picture of a man whose head is tilted to the side and makes a painting where his head is not tilted. You wont be able to do that with imagemagic
I'd be welcome to be proven wrong, but I don't think you'd be able to achieve the results shown here without some sort of stable diffusion or GAN approach.
You can also do much the same with imagemagic, no AI required.