Human-AI joint task performance: Learning from uncertainty in autonomous driving systems

IF 5.7 2区 管理学 Q1 INFORMATION SCIENCE & LIBRARY SCIENCE Information and Organization Pub Date : 2024-01-30 DOI:10.1016/j.infoandorg.2024.100502
Panos Constantinides , Eric Monteiro , Lars Mathiassen
{"title":"Human-AI joint task performance: Learning from uncertainty in autonomous driving systems","authors":"Panos Constantinides ,&nbsp;Eric Monteiro ,&nbsp;Lars Mathiassen","doi":"10.1016/j.infoandorg.2024.100502","DOIUrl":null,"url":null,"abstract":"<div><p>High uncertainty tasks such as making a medical diagnosis, judging a criminal justice case and driving in a big city have a very low margin for error because of the potentially devastating consequences for human lives<em>.</em> In this paper, we focus on how humans learn from uncertainty while performing a high uncertainty task with AI systems. We analyze Tesla's autonomous driving systems (ADS), a type of AI system, drawing on crash investigation reports, published reports on formal simulation tests and YouTube recordings of informal simulation tests by amateur drivers. Our empirical analysis provides insights into how varied levels of uncertainty tolerance have implications for how humans learn from uncertainty in real-time and over time to jointly perform the driving task with Tesla's ADS. Our core contribution is a theoretical model that explains human-AI joint task performance. Specifically, we show that, the interdependencies between different modes of AI use including <em>uncontrolled automation</em>, <em>limited automation</em>, <em>expanded automation</em>, and <em>controlled automation</em> are dynamically shaped through humans' learning from uncertainty. We discuss how humans move between these modes of AI use by increasing, reducing, or reinforcing their uncertainty tolerance. We conclude by discussing implications for the design of AI systems, policy into delegation in joint task performance, as well as the use of data to improve learning from uncertainty.</p></div>","PeriodicalId":47253,"journal":{"name":"Information and Organization","volume":"34 2","pages":"Article 100502"},"PeriodicalIF":5.7000,"publicationDate":"2024-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1471772724000022/pdfft?md5=aa15c890f5d2c514a3e18cf7ce241793&pid=1-s2.0-S1471772724000022-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Information and Organization","FirstCategoryId":"91","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1471772724000022","RegionNum":2,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"INFORMATION SCIENCE & LIBRARY SCIENCE","Score":null,"Total":0}
引用次数: 0

Abstract

High uncertainty tasks such as making a medical diagnosis, judging a criminal justice case and driving in a big city have a very low margin for error because of the potentially devastating consequences for human lives. In this paper, we focus on how humans learn from uncertainty while performing a high uncertainty task with AI systems. We analyze Tesla's autonomous driving systems (ADS), a type of AI system, drawing on crash investigation reports, published reports on formal simulation tests and YouTube recordings of informal simulation tests by amateur drivers. Our empirical analysis provides insights into how varied levels of uncertainty tolerance have implications for how humans learn from uncertainty in real-time and over time to jointly perform the driving task with Tesla's ADS. Our core contribution is a theoretical model that explains human-AI joint task performance. Specifically, we show that, the interdependencies between different modes of AI use including uncontrolled automation, limited automation, expanded automation, and controlled automation are dynamically shaped through humans' learning from uncertainty. We discuss how humans move between these modes of AI use by increasing, reducing, or reinforcing their uncertainty tolerance. We conclude by discussing implications for the design of AI systems, policy into delegation in joint task performance, as well as the use of data to improve learning from uncertainty.

查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
人类与人工智能联合执行任务:从自动驾驶系统的不确定性中学习
高不确定性任务,如医疗诊断、刑事司法案件判决以及在大城市中驾驶,由于可能对人类生命造成毁灭性后果,因此出错的可能性非常低。在本文中,我们将重点研究人类在使用人工智能系统执行高不确定性任务时如何从不确定性中学习。我们分析了特斯拉的自动驾驶系统(ADS),这是人工智能系统的一种类型,我们参考了碰撞调查报告、正式模拟测试的公开报告以及业余驾驶员在 YouTube 上进行非正式模拟测试的录像。我们的实证分析深入揭示了不同程度的不确定性容忍度如何影响人类如何实时并随着时间的推移从不确定性中学习,从而与特斯拉自动驾驶系统共同完成驾驶任务。我们的核心贡献是建立了一个理论模型,用于解释人类与人工智能联合执行任务的情况。具体来说,我们表明,人工智能的不同使用模式(包括不受控自动化、有限自动化、扩展自动化和受控自动化)之间的相互依存关系是通过人类从不确定性中学习而动态形成的。我们将讨论人类如何通过提高、降低或加强其对不确定性的容忍度,在这些人工智能使用模式之间进行转换。最后,我们将讨论对人工智能系统设计、联合任务执行中的授权政策以及利用数据改善从不确定性中学习的影响。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
CiteScore
11.20
自引率
1.60%
发文量
18
期刊介绍: Advances in information and communication technologies are associated with a wide and increasing range of social consequences, which are experienced by individuals, work groups, organizations, interorganizational networks, and societies at large. Information technologies are implicated in all industries and in public as well as private enterprises. Understanding the relationships between information technologies and social organization is an increasingly important and urgent social and scholarly concern in many disciplinary fields.Information and Organization seeks to publish original scholarly articles on the relationships between information technologies and social organization. It seeks a scholarly understanding that is based on empirical research and relevant theory.
期刊最新文献
Stability and change in digital transformation: A repertoire model of institutionally embedded technology affordances Virtual social contagion in online support communities Transformed knowledge work infrastructures in times of forced remote work Editorial Board Digital innovation, platforms, and global strategy
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1