為何應該以人工智能強化倫理衝突的緊急決策？

針對緊急且涉及倫理兩難或衝突而人們難以做出良好決策的事項上，本文主張應該設計人工智能（artificial intelligence，簡稱AI）來改善甚至取代人們做倫理決策。這種AI倫理衝突決策系統，表面上是取代人類做出決策，但實質上是人類借助AI做出較佳的倫理決策，這種決策是從資料驅動到AI驅動。本文透過檢討那些反對以AI做倫理決策的方法論與本體論等兩類批評，並從檢討中指出這兩類批評，有哪部分值得AI開發者警惕，有哪部分則是對於這類倫理決策系統的誤解。本文並進一步指出，這種借助AI來改善涉及公共事務的緊急倫理衝突決策，若能從AI系統設計到部署的每個階段，皆納入決策利益相關人員甚至公眾的參與，便是辛諾特－阿姆斯壯與斯科堡於2021年所主張的「人工改良民主」。

關鍵字

人工智能；倫理決策；緊急情況；倫理兩難／衝突；公共參與及民主

並列摘要

This article argues that artificial intelligence (AI) should be designed to improve or even replace human decision-making in urgent situations involving ethical dilemmas or conflicts, which are well-known to be difficult for people to make good decisions. While the AI ethical conflict decision-making system may seem to supplant human decision-making, in reality, humans utilize AI to make better ethical decisions, transitioning from data-driven to AI-driven decision-making. By examining two types of criticisms-methodology and ontology-against using AI in ethical decision-making, this article points out which parts AI developers should be aware of and which are misunderstandings of these systems. Furthermore, the article suggests that involving decision-makers and even the public in every stage of AI system design and deployment can be viewed as an "Artificial Improved Democracy" proposed by Sinnott-Armstrong and Skorburg in 2021, which could enhance the development of AI ethical decision-making systems involving public affairs.

並列關鍵字

Artificial Intelligence (AI) ； Ethical decision-making ； Urgent situations ； Ethical dilemmas/coflicts ； Public participation and democracy

參考文獻

甘偵蓉、許漢（2020）。〈AI 倫理的兩面性初探：人類研發 AI 的倫理道德與 AI 的倫理規範〉，《歐美研究》季刊，50 卷，2 期，231-292。DOI: 10.7015/JEAS.202006_50(2).0005

陳瑞麟（2020）。〈科技風險與倫理評價：以科技風險倫理來評估台灣基改生物與人工智能的社會爭議〉，《科技、醫療與社會》，30：13-65。DOI: 10.6464/TJSSTM.202004_(30).0001

劉育成（2020）。〈如何成為「人」：缺陷及其經驗作為對人工智能研究之啟發—以自動駕駛技術為例〉，《資訊社會研究》，38：93-126。DOI: 10.29843/JCCIS.202001_(38).0006

劉湘瑤、張震興、張瓅勻、趙恩、李思賢（2021）。〈人工智能倫理的挑戰與反思：文獻分析〉，《資訊社會研究》，41：27-64。DOI: 10.29843/JCCIS.202107_(41).0003

Awad, E., Dsouza, S., Kim, R., Schulz, J., Henrich, J., Shariff, A., Bonnefon, J.-F., Rahwan, I. (2018). The moral machine experiment. Nature, 563, 59–64. Retrieved July 21, 2023, from https://doi.org/10.1038/s41586-018-0637-6

主題瀏覽