Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems 2021
DOI: 10.1145/3411764.3445307
|View full text |Cite
|
Sign up to set email alerts
|

Player-AI Interaction: What Neural Network Games Reveal About AI as Play

Abstract: The advent of artificial intelligence (AI) and machine learning (ML) bring human-AI interaction to the forefront of HCI research. This paper argues that games are an ideal domain for studying and experimenting with how humans interact with AI. Through a systematic

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

0
12
0
1

Year Published

2021
2021
2024
2024

Publication Types

Select...
6
1
1

Relationship

1
7

Authors

Journals

citations
Cited by 32 publications
(13 citation statements)
references
References 80 publications
(75 reference statements)
0
12
0
1
Order By: Relevance
“…This starts from Malone's [32] "Optimal level of informational complexity" and extends to Gentner and Grudin [16], and to Resnick et al 's [37] recommendations that tools provide a "low threshold" to help novice users get started, and be "self-revealing" of more sophisticated capabilities over time. most recently, Zhu et al [44] recommend promotion of discovery-based learning. Our designers value this exploration-based revelation of detail; in session 3, for instance, Gold describes exploring the details underlying default agent behavior, saying "And then if you get curious and you want to see at a lower level, you could keep expanding that, and seeing like, 'Oh, that means that it set all of these variables to these defaults, and it activated these components, and... ' you know under senses, like, 'Oh, hearing!…”
Section: Related Workmentioning
confidence: 99%
See 1 more Smart Citation
“…This starts from Malone's [32] "Optimal level of informational complexity" and extends to Gentner and Grudin [16], and to Resnick et al 's [37] recommendations that tools provide a "low threshold" to help novice users get started, and be "self-revealing" of more sophisticated capabilities over time. most recently, Zhu et al [44] recommend promotion of discovery-based learning. Our designers value this exploration-based revelation of detail; in session 3, for instance, Gold describes exploring the details underlying default agent behavior, saying "And then if you get curious and you want to see at a lower level, you could keep expanding that, and seeing like, 'Oh, that means that it set all of these variables to these defaults, and it activated these components, and... ' you know under senses, like, 'Oh, hearing!…”
Section: Related Workmentioning
confidence: 99%
“…Additionally, we explore the detailed requirements for AI design tools, specifically, and how they relate to gameplay. Zhu et al [44] explore connections between gameplay and AI-human interaction, arguing that "AI as play can expand current notions of human-AI interaction, which are predominantly productivity-based." They suggest using play to help people discover capabilities of an AI-enabled tool.…”
mentioning
confidence: 99%
“…This also entails an understanding of how the introduction of AI-systems will influence traditional decision patterns. Consequently, there has been a lot of behavioral research focusing on how humans make use of or rely algorithmic and AI-systems [42,46,114], or cooperate with them [41,66,117,149,168]. In many bargaining situations, however, people are not users or collaborators, but competitors, or simply the ones affected by the machine's decision.…”
Section: Introductionmentioning
confidence: 99%
“…Understanding the mental models of users has shown to improve the UX of digital applications [27,33,8]. For AIdriven experiences, understanding users' mental models of the AI and designing the UX accordingly can be instrumental to developing human-centered approaches to AI explainability [16] and human-AI interaction in general [30,2,22,43]. A mental model approach goes beyond users' preferences and information needs associated with a given explanation and provides a rich picture of how users comprehend a system.…”
Section: Introductionmentioning
confidence: 99%
“…The central position of this paper is that AI-based games, particularly the player-AI interaction component [43], offer an ideal domain to study the process in which mental models evolve. Games have a long history of getting users to interact with AI in a variety of forms such as non-player characters [38], procedural content generation [37,31], experience management [32,42] and personalized adaptive games [28,36].…”
Section: Introductionmentioning
confidence: 99%