Hypothesis on the Functional Advantages of the Selection-Broadcast Cycle Structure: Global Workspace Theory and Dealing with a Real-Time World
arXiv Preprint Archive May 20, 2025 Junya Nakanishi, Jun Baba, Yuichiro Yoshikawa et al.
Global Workspace Theory, originally a model of human consciousness, can serve as a cognitive architecture for artificial intelligence and robotics operating in dynamic, real-time environments. This paper argues that the theory's Selection-Broadcast Cycle, where information is selected for global availability and then broadcast to the rest of the system, offers three key benefits: dynamic thinking adaptation, experience-based adaptation, and immediate real-time adaptation. Unlike prior work that examined selection and broadcast separately, this work emphasizes their combined cyclic structure. The authors suggest that this framework enables sophisticated decision-making and adaptive performance in unsupervised, dynamic settings, pointing toward new directions for general-purpose AI and robotics systems.