I am becoming more and more convinced that many techy folks near the AI scene saw that SD et alia can create an image that convincingly (and perhaps even indistinguishably) looks like a very nice digital painting, and based on that data point alone are calling artists obsolete.
I've played around with Dall-E a bit and based on trying to create weird ideas like an army of toddlers in plate armor riding corgis into a medieval battle or a bear riding a bicycle pulling a semi truck, I'm fully convinced that it's over-blown.
It can recreate data close to what it has already seen, just like all neural network techniques. It does poorly outside that domain.
