To what extent is the artificial friend welcome in the good human life?

人造朋友在美好的人类生活中受到何种程度的欢迎？

• 批准号: 2602084
• 金额: --
• 依托单位: University of Leeds
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2602084
• 关键词: what; extent; artificial; friend; welcome

The world we live in has predominantly chosen to embrace rapid advances in data-driven technological innovation; this has played a fundamental role in shaping the contemporary social landscape. One key aspect of this change can be observed in the way that we interact with others - from formulating and maintaining relationships solely with organic agents, to growing a dependency and appreciation of virtual and intelligent artificial agents. I seek to critically evaluate the extent to which our relationships with artificial agents (AA's) can aid in human flourishing, by means of addressing the capacity for AA's to exhibit (an ancient conception of) the virtue of friendship. By contributing to a field that unites moral and technological intelligence, my research seeks to establish the extent to which an artificial friend is (or should be) welcome in the good human life. This research thesis will defend and utilise a model that will aid in conceptualising exactly what constitutes an artificial friend. The PGR proposes to adopt the 'Method of Abstraction', or more specifically, epistemological levelism. This methodology involves explicitly specifying observations of a system (the 'object of study'). The finite but non-empty set of observables specified is referred to as the level of abstraction, (loA), and determined in accordance with what set is required for an adequate analysis of that system. For example, in the case of wine, an "olfactory" LoA would incorporate a set of observables such as 'bouquet' 'aroma', and 'citrus' etc., whereas a "purchasing" LoA might include a set of observables such as 'maturity', 'supplier', and 'region' etc. (Floridi, 2008: p.309). For the PGR's thesis, the system the PGR will analyse is 'moral agency'. Floridi and Saunders argue that since human beings are paradigmatic moral agents, the set of observables required to analyse moral agency must be at a human level of abstraction. They argue that this set of observables 'probably' include (Floridi & Saunders, 2004: p. 357-8):a) Autonomy: The capacity to perform internal transitions - i.e. transitions that change its state without direct response to interaction.b) Adaptability: The agents interactions/experiences can change the transition rules by which it changes state (e.g., machine learning). c) Interactivity: The agent and environment can act (take input or yield output) upon each other.When such properties are observed in artificial agents, it is argued that (at that LoA) AA's qualify as moral agents - meaning it is at least theoretically plausible that such entities can exhibit virtuous acts (Floridi and Saunders, 2004: p.349). The focus of my thesis specifically concerns the AA's capacity to exhibit the virtue of friendship. To examine the nature of friendship and its relationship to the good life, the PGR will build upon my master's dissertation (concerning the nature of friendship in Plato's Symposium and Lysis) by expanding the PGR's knowledge to explicate how Platonic conceptions of friendship are related to the good life. Furthermore, the PGR will draw upon Aristotle's conception of the virtue of friendship as presented in Nichamonian Ethics books 8 &9.

我们生活的世界主要选择拥抱数据驱动的技术创新的快速发展；这在塑造当代社会面貌方面发挥了根本作用。这种变化的一个关键方面可以从我们与他人互动的方式中观察到——从仅仅与有机代理建立和维持关系，到越来越依赖和欣赏虚拟和智能的人工代理。我试图批判性地评估我们与人工智能体（AA's）的关系在多大程度上有助于人类的繁荣，通过解决AA's展示（一个古老的概念）友谊美德的能力。通过对一个将道德和技术智能结合在一起的领域做出贡献，我的研究试图确定在美好的人类生活中，人工朋友在多大程度上是（或应该是）受欢迎的。这篇研究论文将捍卫和利用一个模型，这将有助于概念化什么是人造朋友。PGR建议采用“抽象方法”，或者更具体地说，认识论层次主义。这种方法包括明确地指定对系统（“研究对象”）的观察。指定的有限但非空的可观察对象集称为抽象级别（loA），并根据对该系统进行充分分析所需的集合来确定。例如，以葡萄酒为例，“嗅觉”LoA将包括一系列可观察的指标，如“酒香”、“香气”和“柑橘”等，而“采购”LoA可能包括一系列可观察的指标，如“成熟度”、“供应商”和“地区”等（Floridi， 2008：第309页）。对于PGR的论题，PGR将分析的系统是“道德代理”。弗洛里迪和桑德斯认为，既然人类是典型的道德行为人，那么分析道德行为人所需要的一组可观察到的东西就必须处于人类的抽象水平。他们认为，这组可观察的“可能”包括（Floridi & Saunders， 2004：第357-8页）：a)自主性：执行内部转换的能力-即在没有直接响应交互的情况下改变其状态的转换。b)适应性：代理的交互/经验可以改变它改变状态的过渡规则（例如，机器学习）。c)交互性：agent和环境可以相互作用（接受输入或产生输出）。当这种特性在人工行为体中被观察到时，有人认为（在LoA下）AA有资格成为道德行为体——这意味着至少在理论上这些实体可以表现出道德行为是合理的（Floridi和Saunders， 2004：第349页）。我论文的重点是关注AA展示友谊美德的能力。为了研究友谊的本质及其与美好生活的关系，PGR将在我的硕士论文（关于柏拉图的《会饮篇》和《解》中友谊的本质）的基础上，通过扩展PGR的知识来解释柏拉图的友谊概念如何与美好生活相关。此外，PGR将借鉴亚里士多德在《尼迦摩尼伦理学》第8卷和第9卷中提出的友谊美德的概念。