Proceedings of the Annual Meeting of the Cognitive Science Society

DiPaola’s research endeavors to build top down Artificial Intelligence (AI) models of human creativity, empathy and expression for both use in new forms of computation systems as well as analysis of how the creative mind works. In doing so he has interviewed hundreds of artists, writers and musicians on how they perceive their creative talent and its originals. Combined with research from neuro-aesthetics and computer modelling, DiPaola notes that while many creative individuals report that they believe new insights as coming into them from an external source during creative flow, that evidence point to these new creative ideas and interpretations often more likely have internal roots from the individual’s, mid and long term past experiences and processes. DiPaola attempts to model this and other human creativity processes in computational form often as AI systems such as deep learning, reinforcement learning and evolution programming. Two efforts underway in DiPaola’s research lab are mapping out the creative process of a fine art portrait painter using 5 hierarchical AI systems, as well as modelling an empathetic embodied character agent who can understand emotions from those she talks with and construct creative narrative or quote like responses.