{"title":"Unpacking Human and AI Complementarity: Insights from Recent Works","authors":"Yuqing Ren, X. Deng, K. D. Joshi","doi":"10.1145/3614178.3614180","DOIUrl":null,"url":null,"abstract":"In this editorial, we draw insights from recent empirical studies to answer some key questions related to human and AI complementarity. There is consensus regarding the strengths of machine intelligence in performing structured and codifiable tasks and complementing two human limitations: lack of consistency and inability to unlearn conventional wisdom. To work effectively with AI, humans need to possess not only AI skills but also domain expertise, job skills, and metaknowledge to accurately assess human capabilities and AI capabilities. We identify several future directions in understanding the effects of human expertise and experiences on algorithmic appreciation, the mutual learning and adaptions between humans and AI, and the boundary conditions of effective human and AI complementarity.","PeriodicalId":46842,"journal":{"name":"Data Base for Advances in Information Systems","volume":null,"pages":null},"PeriodicalIF":2.8000,"publicationDate":"2023-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Data Base for Advances in Information Systems","FirstCategoryId":"91","ListUrlMain":"https://doi.org/10.1145/3614178.3614180","RegionNum":4,"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
In this editorial, we draw insights from recent empirical studies to answer some key questions related to human and AI complementarity. There is consensus regarding the strengths of machine intelligence in performing structured and codifiable tasks and complementing two human limitations: lack of consistency and inability to unlearn conventional wisdom. To work effectively with AI, humans need to possess not only AI skills but also domain expertise, job skills, and metaknowledge to accurately assess human capabilities and AI capabilities. We identify several future directions in understanding the effects of human expertise and experiences on algorithmic appreciation, the mutual learning and adaptions between humans and AI, and the boundary conditions of effective human and AI complementarity.