This paper proposes a structural-functional framework for defining intelligence at a macro level. Rather than viewing intelligence as a measure of performance, task success, or domain-specific capability, the framework characterises it as a property that emerges from the interaction among information, structure, weighting, and constraint. The central claim is that intelligence is not determined solely by the quantity of information within a system, but by the system's capacity to traverse meaningful pathways through a structured, weighted informational space. The framework is represented by I = C × Φ(S, St, W), where intelligence emerges from the constrained space of traversable inference paths generated within a structured information system. The model is intended as a general framework applicable to biological, artificial, computational, and hypothetical systems. The paper discusses the core definition, structural representation, conceptual advantages, and outlines key open problems, including the formal definition of Φ and its subproblems related to information (S), structure (St), weights (W), constraints (C), and temporal evolution.The framework is also available at: https://github.com/elvisho754-svg/Universal-intelligence-framework
Tsz Shing Ho (Thu,) studied this question.