Two Different Goals: Dissociating the Machine Mind from Artificial Intelligence(两个不同的目标:机器之心与人工智能的概念解绑)

This repository contains the bilingual manuscript and LaTeX sources for a conceptual-engineering paper, the fourth in the human-referential fitting theory program of AGI (AGI 人本拟合论). The paper states and defends the unbundling postulate: artificial intelligence and the machine heart (机器之心) are two different goals — the difference is teleological, not metaphysical — and no transfer of progress is licensed between them. Contributions: (1) a genealogy of the two programs' independent formation and institutionalization, including the 2022–2026 dissociation wave; (2) a four-mechanism diagnosis of the binding (conceptual slippage, narrative structure, commercial incentive, linguistic inertia); (3) the postulate on a teleological axis, distinct from Searle's weak/strong distinction, Block's property distinction, indicator-based assessment frameworks, Seth's epistemological argument, and Pesonen & Reijula's relevance argument, with the progress-transfer illusion as its central corollary; (4) dual-track governance, the full normative responsibility of the engineering goal without a mind verdict, discourse segregation, and rejection clauses stating what findings would re-merge the goals. All citations were verified against primary sources; a precedent-search log documents the absence claim. Bilingual manuscript: English (paper.pdf) and Chinese (paper_zh.pdf), with LaTeX sources; compiled with tectonic.