Opening the moral machine’s cover: How algorithmic aversion shapes autonomous vehicle adoption
通过哥伦比亚实验，研究了自动化程度如何影响算法厌恶（包括道德判断和应急评估），进而影响用户对自动驾驶汽车的接受意愿，发现感知享乐在其中起中介作用。
Autonomous driving technology has made its way into the market at various levels, yet fully autonomous vehicles remain unavailable. The psychological barriers that must be overcome before fully automated vehicles (AVs) become mainstream are numerous. In addition to technological advancements, persuading consumers to transition from the traditional human-driven model to AVs poses a significant challenge. According to the Moral Machine Experiment, Latin American countries form distinct sub-clusters and exhibit the highest preference for action in moral decision-making. To foster user acceptance of AVs in these countries, it is imperative to comprehend cognitive, affective, and ethical factors. To this end, we conducted experiments with respondents from Colombia to examine how varying levels of automation influence algorithm aversion and user acceptance. Algorithm aversion is explored from two perspectives: ethical judgment and behavior, and emergency evaluation and performance. Our findings reveal two key insights. Firstly, higher levels of automation negatively impact people’s assessment of the emergency evaluation capabilities of AVs, partially contributing to algorithm aversion. Secondly, the intention to use AVs is adversely affected by algorithm aversion, encompassing both ethical considerations and emergency performance-related aspects. Furthermore, mediation analysis demonstrates that perceived hedonism elucidates the inverse relationship between algorithm aversion and the intention to use AVs.