信任:个体、团体、人工智能体  

Trust:individuals,groups and artificial intelligence agents

作　　者：甘绍平 GAN Shao-ping

机构地区：[1]贵州大学哲学学院,贵州贵阳550025

出　　处：《东南大学学报(哲学社会科学版)》2025年第2期32-41,146,共11页Journal of Southeast University(Philosophy and Social Science)

摘　　要：信任作为人类存在和人际交往的基本条件,先于道德与法律并与它们并列。信任体现出了一种复杂的结构,“一种介于知与未知的中间状态”。信任最大的存在价值在于减少复杂性,但信任并非总是一种正确的行为方式,与时空情境相适的质疑与监控对于人类生活永远不可或缺。信任发生于具有自我选择能力的行为主体之间,此行为主体不仅指人类个体,而且也是指在功能上拥有自主行为能力的团体与机构。人工智能体并非行为主体,将信任概念用于人工智能系统上仅具有某种拟人化和比喻性的意义。但人工智能体决策过程的可解释性以及与此相关的可信性问题,却构成人工智能研究领域重大的伦理议题。在对信任问题的探究渗透到几乎所有的社会核心领域以及相关的人文社会科学的背景下,信任这一对于哲学边缘性的但日显重大意义的概念,必然成为作为与其他学科处于交叉关系的应用伦理学的研议对象。Trust,as a fundamental condition for human existence and interpersonal interaction,precedes morality and law while coexisting with them.Trust embodies a complex structure—“an epistemologically intermediate state between knowledge and the unknown.”Its greatest existential value lies in reducing complexity,yet trust is not always a correct behavioral mode;skepticism and oversight adapted to spatiotemporal contexts remain indispensable to human life.Trust arises between autonomous agents capable of self-selecting actions,where“agents”refer not only to human individuals but also to groups and institutions endowed with functional autonomy.Artificial intelligence(AI)agents,however,are not autonomous actors.Applying the concept of trust to AI systems carries only anthropomorphic and metaphorical significance.Nonetheless,the explainability of AI decision-making processes and the associated issue of trustworthiness constitute critical ethical challenges in AI research.As inquiries into trust permeate nearly all core social domains and related humanities and social sciences,trust—a concept once marginal in philosophy but increasingly pivotal—will inevitably become a subject of study for applied ethics,which intersects with other disciplines.

关 键 词：信任 个体 团体 人工智能体 可信性 可解释性 

分 类 号：B82-05[哲学宗教—伦理学]