人工智能:技术条件、风险分析和创新模式升级

Artificial Intelligence:Applicability,risk analysis,and innovation mode upgrading

本文从两个方面探讨人工智能(AI)伦理和治理。一是现有AI成果的技术条件——封闭性。现有AI技术的应用要求场景是封闭的,所以AI的强大能力并非完全通过自主学习而获得。本文研究表明,受到广泛关注的用户隐私等问题,源于行业规则滞后、传统设计范式的伦理非封闭性和长期效应的预测能力不足。二是技术应用的外部机制。目前主要依靠熊彼特模式,其先天局限性是诸多伦理风险的深层原因,并限制了新技术在重大社会问题中的应用。本文基于对熊彼特创新向公义创新升级的时代趋势的分析,提出构建公义创新动力机制的若干路径,并以养老产业为例,阐述公义创新在创造性解决重大社会问题中所具有的巨大潜力和效能。

Ethics and governance of Artificial Intelligence(AI) are explored from two aspects in this paper.The first aspect is from the applicability of the existing AI achievements-the closedness,which requires that the application scenarios should be closed and implies that the remarkable capabilities of the existing AI technology is not completely obtained through autonomous learning.This study also shows that the ethical problems which are widely concerned such as that of user privacy are due to the lag of code of practice,the ethical non-closedness in the traditional design paradigm and the insufficient ability to predict long-term effects.The second aspect is from the external mechanism of technology application,Schumpeter’s innovation, whose inherent limitations are the deep cases of many ethical risks and restrict the application of AI technology in major social problems,as revealed in this paper for the first time.Based on the analysis of the trend of upgrading from Schumpeter’s innovation to Gong-Yi innovation,some ways to construct the dynamic mechanism of Gong-Yi innovation is put forth.In addition,a solution to the plight of aged care is proposed to illustrate the tremendous potential and effectiveness of Gong-Yi innovation in creatively solving major social problems.