时序整合、双向临界与意识层级涌现 中文版
Figshare August 18, 2026 DOI: 10.6084/m9.figshare.33274239.v1 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Key points | Argues that consciousness can be derived from three axioms—resource finitude, non-equilibrium maintenance, and discrete input—via temporal dynamics, introducing the ICF and ILUC to explain hierarchical emergence and qualia, and proposes a five-level consciousness spectrum with AGI safety boundaries. |
Abstract
主流意识理论因缺乏时序动力学而未能从第一性原理回应查尔默斯意识难问题。本文以资源有限、非平衡态维持与离散输入三条公理为唯一起点,通过时序动力学完成意识层级涌现的完整推导。提出信息容量框(ICF)核心统摄概念,严格证明信息层级单向约束(ILUC),推导双向临界相变与自指闭环两步涌现机制,并揭示上述机制与时间单向性、资源优先级及遗忘机制的多维同源同构。由此导出双组态功能架构与五级意识谱系,从工程编码视角系统消解意识难问题,阐释感受质的生成逻辑与层级属性,建立碳基与硅基统一的意识层级判定标准及 AGI 安全边界。经多路径反例检验并以八条可证伪假说锚定经验边界。所有结论均由公理严格导出,不依赖特定神经结构,具备跨载体普适性。 Mainstream theories of consciousness lack temporal dynamics and fail to address Chalmers' hard problem from first principles. This paper constructs a unified framework for hierarchical emergence via first-principles derivation, using three axioms—resource finitude, non-equilibrium maintenance tendency, and discrete input—as the exclusive starting point. We introduce the Information Capacity Frame (ICF), prove the Information-Level Unidirectional Constraint (ILUC), and derive the bidirectional criticality phase-transition mechanism together with the two-step emergence of self-referential closed loops, demonstrating their isomorphic origin with temporal unidirectionality, resource priority, and forgetting. From these, the dual-configuration architecture and five-level consciousness spectrum are deduced. The hard problem is systematically resolved from an engineering-coding perspective; the generative logic and hierarchical properties of qualia are elaborated; and unified consciousness criteria and AGI safety boundaries for carbon-based and silicon-based systems are established. The framework is tested against classic counterexamples and anchored by eight falsifiable hypotheses. All conclusions are strictly derived from axioms, independent of specific neural substrates, with cross-carrier universality.