This paper proposes a new theoretical framework for understanding Artificial General Intelligence (AGI) based on a Habitat–Information perspective. We argue that current AI systems are fundamentally limited by their reliance on statistical pattern recognition, lacking true information selection mechanisms and adaptive interaction with external environments. The framework introduces a structured interpretation of intelligence as an evolving system of information processing within dynamic habitats, bridging biological evolution, cognitive systems, and artificial intelligence. This work aims to provide a conceptual foundation for future AGI architectures and interdisciplinary research across AI, cognition, and complex systems.
Zhou Zheng (Thu,) studied this question.