关键词: 亲社会倾向, 感知机器人威胁, 集体焦虑, 外群体, 道德比较

Abstract: Artificial intelligence is bound to pose a threat to humans when it enters human society, and this threat may include concerns about robots replacing human jobs as well as robots blurring the line between humans and machines. This is called the realistic and symbolic threat posed by robots. This robotic threat creates collective angst for human beings, which is a consideration of the future fate of the group and reflects the negative emotions of group members who are worried about the future. When they realize that the group is in crisis, human societies become more murderous and extreme, and their pro-social tendencies are weakened. Of course, this phenomenon only manifests itself for out-group members, while for in-group members the threat of robots does not reduce pro-social tendencies towards them. Moral comparisons also have a moderating effect, and since downward moral comparisons contribute to stronger self-improvement, which creates a moral license to behave morally and at the same time allows one to indulge in immoral behaviors, e.g., a decrease in pro-social behaviors, the effect of robotic threat on pro-social tendencies should only be seen in those who tend to make downward moral comparisons. This research consists of eight recursive studies based on the basic hypothesis that perceived robotic threat reduces people's pro-social tendencies, with collective angst acting as a mediating mechanism, and inside-outside group and moral comparison tendencies moderating its effects. Study 1a and 1b used pre-existing archival databases to calculate the relationship between perceived robot threat at the individual level and pro-social behavior at the national level. Study 2, on the other hand, used a questionnaire to examine collective angst as a mediating variable between robot threat and reduced pro-social tendencies. Study 3 experimentally manipulated robot threat to similarly validate the mediating role of collective angst. Study 4 excluded the influence of other potential mediators. Studies 5 and 6 also used experimental methods to respectively explore the moderating role of internal and external groups and moral comparison tendencies, and to clarify the boundary conditions between perceived robot threat reducing pro-social tendencies. Study 7 further validated the influence of perceived robot threat on prosocial behavior through behavioral observation. The main findings of this research are as follows. First, perceived robotic threat indeed reduces people's pro-social tendencies, even at the group level (Studies 1–7); Second, the mechanism is mediated by collective angst, that is, perceived robotic threat increases collective anxiety, which reduces pro-social tendencies in order to alleviate anxiety (Studies 2–4); Third, This effect is moderated by in-group and out-group distinction, whereas perceived robotic threat primarily reduces pro-social tendencies for out-group members (Study 5); Fourth, this effect is moderated by the tendency to make moral comparisons, that is, perceived robot threat mainly reduces pro-social tendencies for downward moral comparators (Study 6). As findings on the social consequences of robotic threats in prior research have been inconsistent, the results of this research strongly support the conclusion that robotic threats result in negative rather than positive interpersonal outcomes. The reason for this may be the difference between the internal and external groups that people psychologically delineate when they are faced with a specific people to engage in interpersonal behavior after being threatened by a robot. Our research theoretically explains why robotic threats produce negative interpersonal outcomes.

Key words: pro-social tendencies, perceived robot threat, collective angst, out-groups members, moral comparisons