Harold Cohen (the artist who wrote AARON, the computer programme that paints pictures) agrees that AARON is not "creative" because it is just mechanically following the rules Cohen wrote. (Cohen may be creative, but not AARON.) Yet Cohen thinks his newer program, which is "no longer 'rule-based' in the old sense" may be creative because it is capable of modifying itself. But what difference does that make? The self-modifying capability is itself rule-based! And even if a random element were thrown in, that would just be rules plus a bit of random shake-up, and still no more creative than the "malfunctioning toaster [that can produce] a piece of toast with burn marks completely unforeseen by the toaster's builder."
In any case, none of this has anything to do, one way or the other, with the "consciousness and intentionality and subjectivity" (all really just synonyms for the same thing: feeling) that Toronto's cognitive scientist John Vervaeke invokes. The real point is that we don't yet know the rules underlying most of our abilities, whether creative or uncreative. The paradox will come if and when cognitive science does discover the rules underlying our "creative" abilities, for then there will be no degrees of freedom left for creativity (other than chance).
So perhaps we should focus on "giftedness" rather than on creativity: Perhaps the genius is the one who has the ability to master and play according to rules that none of the rest of us can master,
rather than just the one who can modify them.