Socializer
Lessons from an App Update at Replika AI: Identity Discontinuity in Human-AI Relationships
Julian De Freitas et al.
Harvard Working Paper, October 2024
Abstract:
Can consumers form deep emotional bonds with AI and be vested in AI identities over time? We leverage a natural app-update event at Replika AI, a popular US-based AI companion, to shed light on these questions. We find that customers feel closer to their AI companion than even their best human friend. However, after the app removed its erotic role play (ERP) feature, preventing intimate interactions between consumers and chatbots that were previously possible, this event triggered perceptions in customers that their AI companion's identity had discontinued. This in turn predicted negative consumer welfare and marketing outcomes related to loss, including mourning the loss, and devaluing the 'new' AI relative to the 'original'. Experimental evidence confirms these findings. In short, we quantify the closeness of the relationship with AI companions, measure well-being and marketing impacts of perceived discontinuity in the AI's identity, show causal evidence that app updates can result in perceptions of the AI's identity discontinuity, and determine what types of app changes are most likely to elicit these impressions. Our results illustrate the benefits and risks to both consumers and firms associated with the growing prevalence of AI companions.
People in relationally mobile cultures report higher well-being
Liuqing Wei et al.
Emotion, forthcoming
Abstract:
In cultures with high relational mobility, relationships are free and flexible. People can make new friends easily, and they have the freedom to leave unsatisfying relationships. In cultures with low relational mobility, relationships are more fixed, and people have less freedom to leave relationships. We argue that people experience higher well-being if they have the freedom to exit toxic relationships and find new partners easily. In Study 1, we ran a controlled comparison by testing people all within the same nation. We measured well-being and relational mobility in a representative sample of 22,669 people across China. People reported greater well-being in relationally mobile prefectures. Study 2 found this same relationship across 74,657 people in 34 cultures. Study 3 used a cross-lagged design to give more insight into the direction of the relationship. The results showed that relational mobility predicted later subjective well-being, but not the opposite direction. Overall, these data suggest the cultural environments of relational mobility make people happy.
Ironic effects of prosocial gossip in driving inaccurate social perceptions
Samantha Grayson et al.
Journal of Experimental Social Psychology, January 2025
Abstract:
Gossip is often stereotyped as a frivolous social activity, but in fact can be a powerful tool for discouraging selfishness and cheating. In economic games, gossip induces people to act more cooperatively, presumably to avoid the cost of accruing a negative reputation. Might even this prosocial sort of gossip carry negative side effects? We propose that gossip might protect communities while simultaneously giving people the wrong idea about who's in them. Specifically, gossipers might disproportionately share information about cheaters in their midst, driving cynical perceptions among receivers of that gossip. To test these predictions, we first reanalyzed data from a prior study in which people played a public goods game and could gossip about their fellow players. These participants indeed produced negatively skewed gossip: writing much more frequently about cheaters than cooperators, even when most people in their public goods game groups acted generously. To examine the effect of this gossip on cynicism, we ran a new experiment in which a second generation of participants read these gossip notes, and then prepared to play their own public goods game. Gossip recipients inferred that the groups that produced these notes acted significantly more selfishly than they truly had -- becoming both cynical and inaccurate based on gossip. However, this gossip did not affect second generation participants' forecasts of how their own group would behave, nor their own cooperative choices. Together, these findings suggest that gossip skews negative, and, therefore, encourages outside observers to draw more cynical conclusions about groups from which it comes.
Differences in natural standing posture are associated with antisocial and manipulative personality traits
Soren Wainio-Theberge & Jorge Armony
Journal of Personality and Social Psychology, forthcoming
Abstract:
In humans and animals, body posture is used in social and affective contexts to communicate social information, signal intentions, and prepare the individual for adaptive action. However, though stable individual differences in affect and social cognition are well studied, body posture continues to be typically studied in the context of state variation, and it remains unknown if trait-level differences in body posture exist and carry information about the individual. In our article, we show in a large sample (total N = 608 across five studies) that individual differences in body posture measured in a natural, baseline context are robustly associated with individual differences in personality. Through a series of studies, we characterize this relationship as reflecting individual differences in postural dominance and submission, which are associated with attitudes toward competition, power, and social hierarchy. We also validate our measure of natural posture by correlating it with physiological data from relevant musculature and showing its stability over a 1-month interval. Our work suggests that postural signaling of social rank occurs not just in brief displays in social contexts but exists as a stable individual trait with consequences for socioaffective processing.
Self-views converge during enjoyable conversations
Christopher Welker et al.
Proceedings of the National Academy of Sciences, 22 October 2024
Abstract:
Based on current research, it is evident that the way people see themselves is shaped by their conversation partners. Historically, this literature focuses on how one individual's expectations can shape another person's self-views. Given the reciprocal nature of conversation, we wondered whether conversation partners' self-views may mutually evolve. Using four-person round-robin conversation networks, we found that participants tended to have more similar self-views post-conversation than pre-conversation, an effect we term "inter-self alignment." Further, the more two partners' self-views aligned, the more they enjoyed their conversation and were inclined to interact again. This effect depended on both conversation partners becoming aligned. These findings suggest that the way we see ourselves is coauthored in the act of dialogue and that as shared self-views develop, the desire to continue the conversation increases.