The Global Workspace Theory: A Step Towards Artificial General Intelligence

Abstract

Global Workspace Theory (GWT) and Artificial General Intelligence (AGI) are two concepts in cognitive science and Artificial Intelligence, respectively. This paper discusses the possibility of achieving AGI using a deep learning implementation of GWT. The shared latent space for GWT is trained using the latent spaces of the connected deep learning modules. This implementation aims to enhance the performance of specialized models in their specified tasks and achieve more general functions from single-task/specialized modules. The paper also discusses the possible applications of this implementation in healthcare.