Demystifying Algorithmic 'Dystopia', Understanding Artificial Intelligence: An Ethnomethodological Study of Elementary Practices of Reasoning with Alg

揭秘算法“反乌托邦”，理解人工智能：用算法进行推理的基本实践的民族方法学研究

• 批准号: 2273031
• 金额: --
• 依托单位: University of Liverpool
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2019
• 资助国家: 英国
• 起止时间: 2019 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2273031
• 关键词: Demystifying; Algorithmic; Dystopia; Understanding; Artificial

Rooted in ethnomethodology and science fiction studies, this thesis seeks to demystify Artificial Intelligence (AI) (or algorithmic 'dystopia') by revealing mundane methods of practical reasoning with algorithms and code. The methods/craft we use at work/play with AI are often mystified by depictions of AI in the media as dystopic, less-than-human phenomena. This thesis introduces the reader to the "metamodern turn", estrangement with AI as a decaying novum, and how we build worlds; (e)utopias, dystopias and between. Through the works of Isaac Asimov, Alan Turing, Harlan Ellison, Ursula Le Guin and Kazuo Ishiguro, this thesis pairs the reading of science fiction texts with programming with Python including work/play with; text adventure games, videogame bots, virtual environments, worldbuilding as a group study and computer vision/emotion detection. The purpose is to demonstrate that there is no final theory for 'doing' work/play with AI, but rather the findings point to an expanding, evolving, yet decaying set of human methods for making sense (and 'oddkin') with machines. By addressing these studies/tutorials as a collection of methods, this thesis seeks to draw out the mundane, taken-for-granted troubles and repairs programmers perform. The significance of this study is that it provides a kaleidoscopic view of work/play with AI which embodies the disembodied human counterparts that often remain 'behind the scenes' when we talk about AI.

这篇论文植根于民族方法论和科幻小说研究，试图通过揭示用算法和代码进行实际推理的平凡方法来揭开人工智能(AI)(或算法的反乌托邦)的神秘面纱。我们在工作/娱乐中使用的人工智能方法/工艺经常被媒体描述为反乌托邦的、不像人类的现象而感到困惑。这篇论文向读者介绍了“元现代转向”，对作为腐朽新星的人工智能的疏远，以及我们如何构建世界；(E)乌托邦、反乌托邦和两者之间。通过Isaac Asimov、Alan Turing、Harlan Ellison、Ursula Le Guin和Kazuo Ishiguro的作品，本论文将阅读科幻文本与使用Python编程相结合，包括工作/娱乐、文本冒险游戏、视频游戏机器人、虚拟环境、作为小组学习的世界构建和计算机视觉/情绪检测。这样做的目的是为了证明，目前还没有关于用人工智能进行工作/娱乐的最终理论，而是这些发现指向了一套不断扩大、不断发展、但正在衰落的人类用机器来理解(和‘古怪’)的方法。通过将这些研究/教程作为方法的集合来处理，本论文试图引出程序员执行的平凡的、被认为是理所当然的麻烦和修复。这项研究的意义在于，它为人工智能的工作/娱乐提供了一个万花筒般的视角，它体现了当我们谈论人工智能时，往往留在幕后的非实体人类同行。