Creativity is a very vague word, I'm sure we can come up with definitions of it that let humans keep sole domain over it. But breakthroughs often come from combining domains and concepts, very very rarely do we ever jump out of one local maxima into another, and I'm not even convinced that Copernicus counts as that. There's a reason why there are so many examples of the same breakthrough happening in multiple places in the world independently - innovation is a slow gradual collaborative process and not plateaus waiting for men of genius to have a spark of inspiration.

Also I'm not convinced that a computer couldn't have discovered the earth revolves around the sun - it's hard to make machine learning jump out of local maxima, but it does happen, and I can see some hidden layers becoming far more efficient at predicting outcomes by stumbling across a model that centered the sun. That being said - there likely are examples of things that computers couldn't have theoretically figured out the model for, but I'm hard pressed to think of one.