摘要
以人工智能为驱动的人机交互技术已经扩散到各类社会生活场景中,由其引发的信任问题得到了普遍关注。当前的人机信任研究多分散于对不同信任要素的分析上,而机器能动性的切入能够为建构人机交互的信任机制提供有效支撑。人机交互过程中机器能动性的影响要素主要包括自动化程度、不透明度和交互强度,这三个要素共同影响了人机交互中“电子信任”的可信度。因此,可以通过调控机器能动性的方式建构人机交互中的信任机制,具体包括建构可信任自主系统、可解释的“白箱”系统以及信任校准的交互系统。
Human-machine interaction(HMI)technology driven by artificial intelligence has been widely spread to all kinds of social life scenes,and the trust problem caused by it has been widely concerned.The current research on human-machine trust mostly focuses on the analysis of different trust elements,and lack of constructing the human-machine trust mechanism which can be effectively overcome by the introduction of machine agency.The enhancement of machine agency in the process of human-machine interaction is mainly reflected in three elements:automation,opacity and interaction intensity,which affect the credibility of e-trust in human-machine interaction.Therefore,the trust mechanism in human-machine interaction can be constructed by regulating the mode of action of machine agency,including constructing trusted autonomous system,interpretable whitebox system and trust calibrated interactive system.
作者
于雪
YU Xue(Department of Philosophy,Dalian University of Technology,Dalian 116024,China)
出处
《自然辩证法研究》
CSSCI
北大核心
2022年第10期43-49,共7页
Studies in Dialectics of Nature
基金
国家社科基金重大项目“负责任的人工智能及其实践的哲学研究”(21&ZD063)。
关键词
机器能动性
人机交互
信任
自动化
信任校准
machine agency
human-machine interaction
trust
automation
calibrated trust