理解优先的 AI 工程：目的、参考架构与面向人类理解的架构模式
Understanding-First AI Engineering: Purposes, Reference Architecture, and Architectural Patterns for Human Understanding
AI 导读
arXiv 论文提出“理解优先的 AI 工程”，将面向特定目的的人类理解作为指导需求、设计与运行的一等设计原则。论文将理解的目的分为反思性自主、有效行动、集体治理与认识价值四类，给出通过共享记录连接生产、验证、使用与学习探索的参考架构，并提出并行学习、解释性模型、目标探索、受众特定视图与分布式专长五种候选架构模式，通过两个假设用例的定性场景分析展示其应用，但未验证实证有效性。
正文 · 原文
Abstract:AI systems can produce reliable results without developing human understanding. However, human understanding remains valuable for shaping subsequent AI-supported work, forming preferences, governing collective choices and satisfying curiosity, even when human participation is unnecessary for correcting errors. We propose understanding-first AI engineering, which treats purpose-specific human understanding as a first-class design principle guiding requirements, design and operation. The purposes of understanding are organised into four complementary rationales: reflective agency, effective action, collective governance and epistemic value. A reference architecture connects production, verification, use, and learning and exploration through shared records, supporting understanding through explanation and participation. Five candidate architectural patterns guide choices about parallel learning, explanatory models, goal exploration, audience-specific views and distributed expertise. Qualitative scenario analysis of two hypothetical use cases demonstrates how the purpose classification, reference architecture and selected patterns guide investigation, participation and decision processes while retaining operational controls. The cases illustrate design consequences and limitations rather than establish empirical effectiveness. The approach makes human understanding an explicit engineering commitment whose justification need not depend on improving the original AI task.
| Subjects: | Artificial Intelligence (cs.AI); Software Engineering (cs.SE) |
| Cite as: | arXiv:2611.13022 [cs.AI] |
| (or arXiv:2611.13022v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13022 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qinghua Lu [view email]
[v1]
Mon, 9 Nov 2026 07:26:25 UTC (4,392 KB)
来源：arXiv cs.AI · arxiv.org