Can AI be creative? Types of Creativity and How to Cultivate Them in Machines

The integration of artificial intelligence into creative professions has sparked fundamentalquestions about the nature of creativity itself. This paper investigates whether AI systemscan be considered truly creative by examining the technical processes underlying AI-drivencreativity and analyzing how different types of creativity – combinatorial, exploratory, andtransformative – can be encoded and operationalized in computational systems. Througha comprehensive analysis of generative AI architectures, particularly diffusion models, andempirical research on creative outputs, we demonstrate that while AI excels at combinato-rial and exploratory creativity within well-defined conceptual spaces, it fundamentally lacksthe intentionality, subjective experience, and consciousness that characterize authentic hu-man creativity. Our findings suggest that the most valuable creative outcomes emerge fromhuman-AI collaboration rather than autonomous machine generation, necessitating a recon-ceptualization of creativity as a spectrum rather than a binary human exclusive capacity.