In the coming years, human-AI cooperation will become an even more central part of many people’s private and work lives, supporting and shaping cognitive processes, while also influencing social dynamics. To understand the conditions under which humans choose to cooperate with AI, researchers have widely utilized evolutionary theories and game-theoretic approaches. However, these frameworks primarily emphasize utility maximization and strategic behavior, overlooking the subjective, experiential dimension of cooperation. To overcome this limitation, we here propose an integrated framework for the emergence of human-AI cooperation, which combines a mechanistic layer drawn from evolutionary theories of cooperation with an experiential layer provided by self-determination theory. We further propose to operationalize the human cooperation intent as the perceived balance between benefits and costs of cooperating, which is moderated by the extent to which psychological needs are satisfied through the cooperation. Our framework offers a novel approach for experimentally testing the formation a human-AI cooperation intent, highlighting not only when and why cooperation may occur, but also how it can be designed to be intrinsically motivating and meaningful for users.

This workshop theme centers on the development of AI agents and systems that are capable of understanding, adapting to, and reacting to collaborate with humans in compliance with the social norms. These systems leverage insights from social psychology, cognitive science, robotics, and AI to interpret social cues, anticipate the needs of others, and coordinate actions effectively within dynamic and often unpredictable contexts. We focus on embedding social awareness into AI systems, leading to Cooperative Intelligence [23] which focuses on building trust and relationship between humans and intelligent systems, instead of focusing on functions to replace humans. This paradigm is expected to realize a hybrid society, where humans coexist with ubiquitous intelligent agents.

Activity trackers are promising tools to increase the motivation to be physically active, thus strengthening users’ physical constitution and potentially preventing cardiovascular diseases. To establish behavioral changes with such beneficial consequences, prolonged continuous use of trackers seems to be necessary. However, the usage behavior of many tracker users is characterized by interruptions or complete discontinuation after only a few months. Which factors determine individual usage trajectories is still unclear. This research sheds light on user diversity to investigate how inter-individual differences are related to reasons for usage interruptions and permanence of abandonment. Results based on a survey of N = 159 former users revealed that usage motives regarding self-determination theory, domain-specific personality traits (affinity for technology interaction), and interaction variables (dependency effect, trust in activity tracker measurement) were related to specific reasons for usage interruptions. Moreover, highly autonomous usage motivation and high trust were linked to more fragile abandonment decisions.

Successful coping with technology is relevant for mastering daily life. Based on related conceptions, we propose affinity for technology interaction (ATI), defined as the tendency to actively engage in intensive technology interaction, as a key personal resource for coping with technology. We present the 9-item ATI scale, an economical unidimensional scale that assesses ATI as an interaction style rooted in the construct need for cognition (NFC). Results of multiple studies (n > 1500) showed that the scale achieves good to excellent reliability, exhibits expected moderate to high correlations with geekism, technology enthusiasm, NFC, self-reported success in technical problem-solving and technical system learning success, and also with usage of technical systems. Further, correlations of ATI with the Big Five personality dimensions were weak at most. Based on the results, the ATI scale appears to be a promising tool for research applications such as the characterization of user diversity in system usability tests and the construction of general models of user-technology interaction.