📄 中文摘要
智能惯性这一概念及其基础物理原理被提出，以量化智能的计算负担。尽管兰道尔原理为信息擦除提供了热力学的基本底线，费舍尔信息则为参数空间中的局部曲率提供了度量，但这些经典框架在稀疏规则约束的领域中仅作为近似有效。它们未能解释在重新配置先进智能系统时，保持符号可解释性所需的超线性、甚至爆炸性的计算和能量成本。智能惯性的引入为理解这一现象提供了新的视角，强调了在复杂智能系统中，计算和能量消耗的深层次物理机制。
智能惯性：物理原理及其应用
出处: Intelligence Inertia: Physical Principles and Applications
发布: 2026年3月25日
📄 English Summary
Intelligence Inertia: Physical Principles and Applications
The concept of intelligence inertia and its underlying physical principles are introduced to quantify the computational burden of intelligence. While Landauer's principle provides a fundamental thermodynamic floor for information erasure and Fisher Information offers a metric for local curvature in parameter space, these classical frameworks serve only as effective approximations within regimes of sparse rule-constraints. They fail to account for the super-linear and often explosive computational and energy costs incurred when maintaining symbolic interpretability during the reconfiguration of advanced intelligent systems. The introduction of intelligence inertia offers a new perspective for understanding this phenomenon, emphasizing the deeper physical mechanisms behind computational and energy expenditures in complex intelligent systems.