A few of the articles about this model make the point that no names (whether a real person's or a brand's) were included in the training data, which makes sense

A few of the articles about this model make the point that no names (whether a real person's or a brand's) were included in the training data, which makes sense, but from a creative point of view there are a lot of other names (cities, monuments, etc.) worth using, whether it is "1970s NYC bar" or "Stonehenge gift shop" or "Everest expedition". And the example prompts included are similarly devoid of stylistic direction beyond the atmospheric or mood. I mean, okay, no "Greg Rutkowski" or "Anne Geddes" or even "Pablo Picasso", but does it understand "Cubism" or "Film Noir" or "Googie"?