把偏见当镜子，让AI设计更公平透明。
Equity Bias: An Ethical Framework for AI Design
- 将偏见视为知识来源的反映，而非需消除的错误
- 提出三阶段生命周期方法，贯穿AI开发全过程
- 适合关注伦理、公平性的开发者与政策制定者
Equity Bias 是一个哲学与实践相结合的框架，旨在构建更智能、更公平的AI系统。基于解释学哲学和认识论不公理论，该框架将偏见视为系统中编码了谁的知识的体现，而非需要消除的错误。与传统方法不同，Equity Bias 不试图消除偏见，而是使其透明且可争议。通过这一过程，它拓展了影响AI设计的视角，并提供了一个将AI系统理解为解释性代理的视角。框架提出了三阶段的AI生命周期方法：'公平考古学'（映射知识与假设）、'共同创造意义'（参与式设计）和'持续问责'（持续评估）。该框架引导开发者、研究人员和政策制定者走向在伦理上可问责、能应对复杂现实挑战的AI。
原文摘要 · Abstract (English)
Equity Bias is a philosophical and practical framework for building smarter, more equitable AI systems. Grounded in hermeneutic philosophy and epistemic injustice theory, it treats bias not as an error to eliminate but as a reflection of whose knowledge is encoded into systems. While traditional approaches aim to reduce or remove bias, Equity Bias instead makes bias transparent and contestable. In doing so, it broadens whose perspectives shape AI and provides a lens for understanding AI systems as interpretive agents. The framework introduces a three-phase AI Life Cycle methodology: 'Equity Archaeology' (mapping knowledge and assumptions), 'Co-Creating Meaning' (participatory design), and 'Ongoing Accountability' (continuous evaluation). Equity Bias guides developers, researchers, and policymakers towards AI that is ethically accountable and capable of addressing complex real-world challenges.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。