人工智能伦理体系:基础架构与关键问题

全球范围内针对人工智能伦理准则的讨论已达成基本共识。在此基础上,本文进一步研究4个关键问题:人工智能伦理体系的运行机制问题、人工智能伦理准则的场景落地问题、人工智能伦理风险的预测判别问题,以及人工智能伦理对重大社会问题综合创新的支撑机制问题。这些问题超越了人工智能伦理准则的范围,却是一种完整、有效的人工智能伦理体系所必须解答的。本文的主要贡献是对这4个问题提出一套建议方案。

Ethical system of artificial intelligence: infrastructure and key issues

On the basis of consensus on ethical guidelines or principles so far reached in the world, four key issues on ethical system of AI will be studied further in this paper:operating mechanisms of AI ethics system, the grounding of AI ethical guidelines in real-world scenarios, prediction and identification of ethical risks of AI technology, and comprehensive innovation mechanisms of handling major social problems supported by the AI ethics system. These issues go beyond the realm of AI ethical guidelines, while they must be solved in order to develop a complete and effective operating system of AI ethics. A preliminary plan for solving these issues is proposed as the main contribution of this paper.