Abstract for: Artificial Creative Intelligence: From Computing to Physiology and Back Again

Undertaking the challenge of achieving artificial creative intelligence (ACI) requires fundamentally rethinking artificial intelligence (AI) as approached with the neural network paradigm. Today’s advances in artificial intelligence rely on this 150-year-old theory. It is time to take one step back and two steps forward in pursuing ACI. Systems thinking approaches artificial creativity by holistically investigating the biology of creativity. What happens in the human body during the creative process? The Delphi method and a series of Gedanken experiments were the primary means for deriving the described ACI model. Experts included in the Delphi approach came from several domains: computer science, physiology, philosophy, and system dynamics, to name a few. Causal loop diagrams were used extensively to drive thinking, pushing out the boundaries of several fields. To date, causal loop diagrams have been developed to model a hybrid human-machine artificial creativity system. Underlying hypotheses and research questions have been posed, such as “Is creativity triggered by an internal chemical imbalance?” and “Is creative homeostasis the absence of creative urges?” The work has identified the need for a multi-disciplinary approach to developing an ACI solution. Neural networks are a 19th-century theory on which the latest AI advances are built, namely, large language models that form the foundation of generative pre-trained transformers. The neural network takes AI only to the point of imitation and not true creativity. A hybrid human-machine creativity system, modeled as a causal loop diagram, is proposed to solve an artificial intelligence problem. System dynamics is being used to solve an AI problem.