Here's the way I understand whether or not artificial intelligence can be creative. As you know, a computer defeated Gary Kasparov, the world's best human chess player. Also, a computer defeated the best Go player.
The common wisdom is that defeating a human at chess just required a computer with enough cycles to try every possible chess move, given the board position at that time. The solution - using enough brute force CPU cycles - is not considered creative because there are a finite number of chess moves, given a position.
Go is another matter. Go requires learning and what one might call "intuition" if it were attributed to a human. In the case of Go, the Google Deep Mind team used neural network programming, not to do a brute force solution, but to program the computer to come up with a better strategy, basically learning as the game progressed. Go is one of the world's mot complicated games and requires adaptability.
One could easily posit this as intelligence, something we attribute uniquely to human beings. However, if we execute the Turing test we could not differentiate between the human and the computer's reasoning powers. In that sense, yes, the program is indeed showing intelligence, not artificial, but the real kind.
I think the most important aspect is the processing power.(with enough memory) What makes us unique in terms intelligence is our immence processing power.