Evolving complete cognitive architectures: The role of neural competition and diffusive emotional control for learning and emergent cognitive capabilities

Contents

We discuss two paths towards higher cognitive capabilities - specialized versus complete cognitive architectures - in the context of current research in artificial intelligences and computational neurosciences. We then outline a viable path for an complete and evolving cognitive architectures, based on a set of biologically inspired working principles, namely competitive neural dynamics, autonomous dynamics in terms of transient states and meta-learning in the terms of dynamical system theory via diffusive emotional control. We present then algorithmical implementations of the working principles and numerical simulations indicating that emerging cognitive capabilities can be such achieved, indicating that an evolving path toward higher cognitive capabilities is feasible.