Human API: A Framework for Cultural Interface Layers in AI-Mediated Knowledge Systems

This paper introduces the Human API, a conceptual framework describing the role of human creators as deliberate interface layers between cultural meaning and machine interpretation in AI-mediated environments. As artificial intelligence systems increasingly mediate access to knowledge through automated retrieval and summarisation — a phenomenon described as the Zero-Click Internet — risks of semantic erosion, authorship displacement, and context loss have emerged. The Human API framework names the strategies creators use to preserve meaning and attribution within these systems. Central to the framework is the concept of the Semantic Exoskeleton: a structured layer of canonical definitions, ontology placement, and machine-readable metadata that stabilises meaning during AI interpretation. The paper formalises these ideas through the Human API Definition Standard (HADS) Version 1.0, establishing a canonical definition, conceptual architecture, implementation taxonomy, and research agenda. Originating within the artistic practice of Metadata Expressionism and AI-Critical Art, the Human API concept has broader applications across cultural production, knowledge work, journalism, and institutional knowledge systems. The framework contributes to scholarship in human–AI interaction, digital humanities, media theory, and algorithmic governance.本文介绍了人类接口（Human API）——一种概念性框架，用于阐述在人工智能介导的环境中，人类创作者作为文化意义与机器解读之间的刻意交互界面层所发挥的作用。 随着人工智能系统愈发多地通过自动化检索与摘要处理来介导知识获取——这一现象被称为零点击互联网（Zero-Click Internet）——语义消解、作者权被取代以及语境流失等风险已然显现。人类接口框架界定了创作者在这类系统中用以保留意义与署名归因的各类策略。 该框架的核心是语义外骨骼（Semantic Exoskeleton）这一概念：它是一套由标准定义、本体论布局以及机器可读元数据构成的结构化层级，可在人工智能解读过程中稳定意义表达。本文通过人类接口定义标准（Human API Definition Standard, HADS）1.0版对上述理念进行了形式化规范，确立了标准定义、概念架构、实现分类法以及研究议程。 人类接口这一概念起源于元数据表现主义（Metadata Expressionism）与批判人工智能艺术（AI-Critical Art）的艺术实践，其应用范畴可拓展至文化生产、知识工作、新闻业以及机构知识系统等多个领域。 该框架为人类-人工智能交互、数字人文、媒介理论以及算法治理领域的学术研究贡献了学术价值。