重构推荐系统，让技术尊重人的尊严与自主
Beyond Algorethics: Addressing the Ethical and Anthropological Challenges of AI Recommender Systems
- 提出三维人本框架，融合政策、跨学科研究与数字素养教育
- 指出现有算法伦理无法应对系统对人性的简化与操控
- 适合关注算法治理、数字伦理与社会影响的研究者
本文探讨人工智能推荐系统（RSs）带来的伦理与人类学挑战。这些系统通过个性化内容推送，不仅反映用户偏好，更主动构建社交、娱乐和电商中的数字体验。其广泛影响引发隐私、自主权与心理健康的担忧。现有“算法伦理”（algorethics）方法仍显不足：系统将人类复杂性简化为可量化数据，利用用户脆弱性，优先追求参与度而非福祉。本文提出一个三维人本推荐系统框架，整合政策与监管、跨学科研究、教育三方面，三者互为支撑——研究提供政策依据，政策建立保护机制与标准，教育提升公众批判性认知。通过连接伦理反思、治理与数字素养，论文主张推荐系统可转向增强人的自主与尊严，而非削弱。
原文摘要 · Abstract (English)
This paper examines the ethical and anthropological challenges posed by AI-driven recommender systems (RSs), which increasingly shape digital environments and social interactions. By curating personalized content, RSs do not merely reflect user preferences but actively construct experiences across social media, entertainment platforms, and e-commerce. Their influence raises concerns over privacy, autonomy, and mental well-being, while existing approaches such as "algorethics" - the effort to embed ethical principles into algorithmic design - remain insufficient. RSs inherently reduce human complexity to quantifiable profiles, exploit user vulnerabilities, and prioritize engagement over well-being. The paper advances a three-dimensional framework for human-centered RSs, integrating policies and regulation, interdisciplinary research, and education. These strategies are mutually reinforcing: research provides evidence for policy, policy enables safeguards and standards, and education equips users to engage critically. By connecting ethical reflection with governance and digital literacy, the paper argues that RSs can be reoriented to enhance autonomy and dignity rather than undermine them.
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。