Social Sciences in China (Chinese Edition)

No. 9, 2025

Machine Consciousness and the Causal Self-Model

(Abstract)

Wu Xiaoan

Inspired by the theoretical virtues of Structural Causal Models (SCM) and drawing on leading contemporary theories of consciousness—including Global Workspace Theory (GWT) and Integrated Information Theory (IIT)—together with advances in artificial intelligence and cognitive science, one can construct a theoretical framework for machine consciousness centered on a causal self-model. Machine consciousness relies not only on general capacities for causal inference; more crucially, the system must internally build a causal model centered on itself. Such a model allows the system to actively identify and understand the effects of its own actions on the environment and the feedback of environmental changes on its own states. Only when the system perceives itself as an active node within the causal chain and can, on this basis, conduct counterfactual simulation and introspection, does it acquire the structural foundation for genuine machine consciousness. This framework offers both rigorous philosophical guidance and a feasible technical pathway for building artificial intelligence systems endowed with genuine consciousness and semantic understanding.