人工智能的监管挑战与路径构建  

The Regulatory Challenge and Path Construction of Artificial Intelligence

作　　者：孙唯特[1] SUN Wei-te(Nanjing University of the Arts,Nanjing 210000,Jiangsu)

机构地区：[1]南京艺术学院,江苏南京210000

出　　处：《电脑与电信》2018年第11期81-85,共5页Computer & Telecommunication

摘　　要：人工智能有着巨大的深度学习和情感伦理识别的潜力,这给人类带来了根本性的生存危机和颠覆性的冲击。然而,监管技术匮乏、监管法律、机制滞后等问题,都无法有效应对未来人工智能的快速发展。因此,本文从传统的行为监管和技术监管等传统防御性监管的维度之外增加以人工智能的自身监管(智能治理)。同时,在技术层面,让人工智能能自主识别情感,建立公平、透明、问责的"程序伦理"算法监管;在监管的引导下,建立统一的行业标准和自身监管系统。这样才能有效地回应人工智能对内含的安全风险及其引发的监管挑战,消除人类的发展顾虑。Artificial intelligence has enormous potential for deep learning and emotional ethical recognition,which has brought about a fundamental crisis of survival and subversive shocks to human beings.However,due to the lack of regulatory technology,regulatory laws,and lags in mechanisms,it is impossible to cope with the rapid development of artificial intelligence in the future.Therefore,this paper adds self-regulation(intelligent governance)of artificial intelligence to the traditional defensive supervision of behavioral regulation and technical supervision.At the same time,at the technical level,artificial intelligence should be able to identify emotions autonomously and establish a fair,transparent,accountable“procedural ethics”algorithm supervision.Under the guidance of supervision,a unified industry standard and its own supervision system should be established.In this way,the security risks and the regulatory challenges caused by the artificial intelligence can be handled properly and human development concerns can be eliminated.

关 键 词：人工智能 监管挑战 智能治理 路径 

分 类 号：D920.0[政治法律—法学]