arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2109.12480cs.HCcs.AI
收藏

Explainability Pitfalls:可解释AI中的暗模式之外

Explainability Pitfalls: Beyond Dark Patterns in Explainable AI

Georgia Institute of Technology(佐治亚理工学院)

Upol Ehsan, Mark O. Riedl

摘要:

为了使可解释人工智能(XAI)系统值得信赖,了解有害影响与产生精心设计的解释同样重要。在本文中,我们解决了一个重要的,但未明确的类型的负面影响XAI。我们引入了可解释性陷阱(EP),这是人工智能解释的意想不到的负面下游效应,即使没有意图操纵用户。EP与黑暗模式不同,但与之相关,黑暗模式是故意欺骗的做法。我们阐明了EP的概念,将其与暗模式区分开来,并强调了陷阱周围的不确定性所带来的挑战。我们使用一个案例研究来解释和操作这个概念,这个案例研究展示了尽管有最好的意图,但毫无疑问的负面影响,例如对数字解释的不必要的信任。我们提出了积极主动和预防性的战略,以解决EP在三个相互关联的层面:研究,设计和组织。

英文摘要:

To make Explainable AI (XAI) systems trustworthy, understanding harmful effects is just as important as producing well-designed explanations. In this paper, we address an important yet unarticulated type of negative effect in XAI. We introduce explainability pitfalls(EPs), unanticipated negative downstream effects from AI explanations manifesting even when there is no intention to manipulate users. EPs are different from, yet related to, dark patterns, which are intentionally deceptive practices. We articulate the concept of EPs by demarcating it from dark patterns and highlighting the challenges arising from uncertainties around pitfalls. We situate and operationalize the concept using a case study that showcases how, despite best intentions, unsuspecting negative effects such as unwarranted trust in numerical explanations can emerge. We propose proactive and preventative strategies to address EPs at three interconnected levels: research, design, and organizational.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑