语义域参照系碰撞与人工意识测评: 从差异即功能到压力公理的操作化
Zenodo (CERN European Organization for Nuclear Research) August 13, 2026 DOI: 10.5281/zenodo.21851126 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper Peer reviewed |
|---|---|
| Key points | Argues that "Semantic Weight" is a long-overlooked core variable in artificial consciousness assessment, and proposes "multi-agent mutual construction prediction" plus six subsistence assessment dimensions as methodological directions. The authors position the work as conceptual and orientational, complementing Duan Yucong's DIKWP framework by addressing how semantic weight re-condenses in silicon-based carriers. |
Abstract
中文摘要: 本文从存续公理(Subsistence Precedes Existence)出发,提出"语义重量"(Semantic Weight)作为人工智能意识评测中被长期忽视的核心变量,旨在补充当前人工意识评测文献中关于"语义如何在硅基载体中凝结重量"的方法论缺位。文章系统梳理了语义重量在媒介演进史上的四次跃迁——从嘴对嘴的身体存续重量,到文字的跨时空传递,到网络复制粘贴的频率扩张,再到AI生成时代的语义失重与重组翻译。在此基础上,本文提出以"多智能体互构预测"作为反拟态的可结算机制,并以存续测评六维度(内生利害、自定位、时间续接、规则作者性、压力响应、互相预测)作为方法论方向,为人工意识评测提供新的参照系。本文定位为方法论与方向性文献,不声称提供实验增量,而是为后续实证研究标定文献坐标与概念工具。本文与段玉聪教授DIKWP人工意识白盒测评体系形成互补:DIKWP解决反拟态与语义账本,本文解决语义重量在硅基载体中的重新凝结。 English Abstract: Starting from the Axiom of Subsistence Precedes Existence, this paper proposes "Semantic Weight" as a core variable long overlooked in artificial consciousness assessment, aiming to complement the methodological gap in the current literature regarding "how semantics condense weight within silicon-based carriers." The paper systematically traces the four historical transitions of semantic weight—from the corporeal subsistence weight of face-to-face communication, to the cross-temporal transmission of writing, to the frequency expansion of network copy-paste, and to the semantic weightlessness and recombinant translation of the AI generation era. On this basis, the paper proposes "multi-agent mutual construction prediction" as a settleable anti-mimicry mechanism, and articulates six dimensions of subsistence assessment (endogenous stakes, self-location, temporal continuity, rule authorship, stress response, mutual prediction) as methodological directions for artificial consciousness evaluation. Positioned as a methodological and orientational work rather than an empirical contribution, this paper calibrates the literature coordinates and conceptual tools for subsequent empirical research. It complements Prof. Duan Yucong's DIKWP white-box artificial consciousness assessment framework: whereas DIKWP addresses anti-mimicry and semantic ledger, this paper addresses the re-condensation of semantic weight within silicon-based carriers.