CHS: Medium: Collaborative Research: Charting a Research Agenda in Artificial Intelligence-Mediated Communication

CHS：媒介：协作研究：制定人工智能介导的沟通研究议程

• 批准号: 1901151
• 负责人: Mor Naaman
• 金额: $ 80.01万
• 依托单位: Cornell University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-01 至 2024-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1901151&HistoricalAwards=false
• 关键词: CHS; Medium; Collaborative; Research; Charting

Artificial Intelligence (AI) algorithms are increasingly augmenting interpersonal communication. What used to be Computer-Mediated Communication (CMC) increasingly involves AI-Mediated Communication (AI-MC): interpersonal communication not simply transmitted by technology but augmented --- or even generated --- by algorithms to achieve specific outcomes. While some simple forms of AI-MC are already prevalent, recent advances in Natural Language Processing provide new directions for augmenting communication online by, for example, modifying texts to include more formal language or enhancing resumes to make them more professional. The advances are not limited to text: increasingly, photos and videos can be automatically manipulated with AI, leading to deep fakes, in which people are shown to act or behave in ways that they never did. Indeed, if a communication is mediated, AI can potentially modify, augment, or even generate the message. AI-MC is therefore likely to have a profound effect on how we communicate, greatly complicating our understanding of technology-mediated human interactions. The project will inform the development of systems that implement AI-mediated communication in a socially desirable and ethically responsible manner. The technical objectives of this project are to develop a framework that charts how AI-MC will impact cyber-human systems research and inform the design of AI-MC technologies. The project will provide some of the first investigations in key areas of AI-MC: 1) the design and perception of AI-MC systems; 2) the potential impact of AI-MC on communication dynamics; 3) the impact of AI-MC on social-psychological dynamics of online self-presentation, with a focus on impression formation and trust, including malicious contexts; and 4) understanding ethical concerns and opportunities around issues like bias, manipulation, and transparency in AI-MC technologies. These objectives will be accomplished through a series of novel empirical studies employing approaches including computational, behavioral, and qualitative methods. These activities will include online and lab experiments that examine behavioral processes and outcomes associated with various forms of AI-MC, the design and development of an AI-MC research platform, as well as a qualitative study of developers, engineers, and designers working with AI-MC systems. In addition, this project will build on the public attention and intrigue around AI to offer design and technology workshops to K-12 students in New York City public schools, using AI-MC to connect the ideas of AI to technologies students use in their everyday lives.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

人工智能(AI)算法正在日益增强人际交流。过去的计算机中介沟通(CMC)越来越多地涉及人工智能中介沟通(AI-MC)：人际沟通不是简单地由技术传递，而是由算法增强--甚至产生--以实现特定的结果。虽然一些简单的AI-MC形式已经很普遍，但自然语言处理的最新进展为加强在线交流提供了新的方向，例如，通过修改文本以包括更正式的语言或增强简历以使其更专业。这些进步并不局限于文本：越来越多的照片和视频可以通过人工智能自动处理，导致深度伪造，在这种情况下，人们会被展示出他们从未做过的行为。事实上，如果沟通是中介的，人工智能可能会修改、增强甚至生成消息。因此，AI-MC可能会对我们的交流方式产生深远的影响，极大地复杂化我们对技术中介的人类互动的理解。该项目将为开发以社会合意和道德负责的方式实施人工智能中介交流的系统提供信息。该项目的技术目标是开发一个框架，显示AI-MC将如何影响网络人类系统研究，并为AI-MC技术的设计提供信息。该项目将在AI-MC的关键领域提供一些初步的调查：1)AI-MC系统的设计和感知；2)AI-MC对交流动态的潜在影响；3)AI-MC对在线自我展示的社会心理动态的影响，重点是印象的形成和信任，包括恶意环境；以及4)了解围绕AI-MC技术中的偏见、操纵和透明度等问题的伦理关切和机会。这些目标将通过一系列新颖的实证研究来实现，这些方法包括计算方法、行为方法和定性方法。这些活动将包括在线和实验室实验，检查与各种形式的AI-MC相关的行为过程和结果，设计和开发AI-MC研究平台，以及对使用AI-MC系统的开发人员、工程师和设计人员进行定性研究。此外，该项目将基于公众对人工智能的关注和兴趣，为纽约市公立学校的K-12学生提供设计和技术研讨会，使用AI-MC将人工智能的想法与学生在日常生活中使用的技术联系起来。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Methodological Middle Spaces: Addressing the Need for Methodological Innovation to Achieve Simultaneous Realism, Control, and Scalability in Experimental Studies of AI-Mediated Communication

方法论的中间空间：满足方法论创新的需求，以在人工智能介导的通信的实验研究中同时实现现实性、控制性和可扩展性

• DOI: 10.1145/3579506
• 发表时间: 2023
• 期刊: Proceedings of the ACM on Human-Computer Interaction
• 作者: Aghajari, Zhila;Baumer, Eric P.;Hohenstein, Jess;Jung, Malte F.;DiFranzo, Dominic
• 通讯作者: DiFranzo, Dominic

AI Writing Assistants Influence Topic Choice in Self-Presentation

• DOI: 10.1145/3544549.3585893
• 发表时间: 2023-04
• 期刊: Extended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems
• 作者: Ritika Poddar;R. Sinha;Mor Naaman;Maurice Jakesch

Computer Vision and Conflicting Values: Describing People with Automated Alt Text.

计算机视觉和冲突的价值观：用自动替代文本描述人。

• DOI: 10.1145/3461702.3462620
• 发表时间: 2021
• 期刊: and Society (AIES
• 作者: Hanley, M.

AI as a moral crumple zone: The effects of AI-mediated communication on attribution and trust

• DOI: 10.1016/j.chb.2019.106190
• 发表时间: 2020-05-01
• 期刊: COMPUTERS IN HUMAN BEHAVIOR
• 影响因子: 9.9
• 作者: Hohenstein, Jess;Jung, Malte
• 通讯作者: Jung, Malte

Negotiating Dyadic Interactions through the Lens of Augmented Reality Glasses

通过增强现实眼镜的镜头来协商二元交互

• DOI: 10.1145/3563657.3595967
• 发表时间: 2023
• 期刊: Designing Interactive Systems Conference (DIS ’23
• 作者: Chung, Ji Won;Fu, Xiyu Jenny;Deocadiz-Smith, Zachary;Jung, Malte F;Huang, Jeff
• 通讯作者: Huang, Jeff