Hirokazu Shirado

Large language models demonstrate strong problem-solving abilities through reasoning techniques such as chain-of-thought prompting and reflection. However, it remains unclear whether these reasoning capabilities extend to a form of social intelligence: making effective decisions in cooperative contexts. We examine this question using economic games that simulate social dilemmas. First, we apply chain-of-thought and reflection prompting to GPT-4o in a Public Goods Game. We then evaluate multiple off-the-shelf models across six cooperation and punishment games, comparing those with and without explicit reasoning mechanisms. We find that reasoning models consistently reduce cooperation and norm enforcement, favoring individual rationality. In repeated interactions, groups with more reasoning agents exhibit lower collective gains. These behaviors mirror human patterns of “spontaneous giving and calculated greed.” Our findings underscore the need for LLM architectures that incorporate social intelligence alongside reasoning, to help address-rather than reinforce-the challenges of collective action.

Virtual reality technologies that enhance realism and artificial intelligence (AI) systems that assist human behavior are increasingly interwoven in social applications. However, how these technologies might jointly influence interpersonal coordination remains unclear. We conducted an experiment with 240 participants in 120 pairs who interacted through remote-controlled robot cars in a physical space or virtual cars in a digital space, with or without autosteering assistance, using the chicken game, an established model of interpersonal coordination. We find that both realism and AI assistance help improve user performance but through opposing mechanisms. Real-world contexts enhanced communication, fostering reciprocal actions and collective benefits. In contrast, autosteering assistance diminished the need for interpersonal coordination, shifting participants’ focus towards self-interest. Notably, when combined, the egocentric effects of autosteering assistance outweighed the prosocial effects of realism. The design of HCI systems that involve social coordination will, we believe, need to take such effects into account.

Forms of both simple and complex machine intelligence are increasingly acting within human groups in order to affect collective outcomes. Considering the nature of collective action problems, however, such involvement could paradoxically and unintentionally suppress existing beneficial social norms in humans, such as those involving cooperation. Here, we test theoretical predictions about such an effect using a unique cyber-physical lab experiment where online participants (N = 300 in 150 dyads) drive robotic vehicles remotely in a coordination game. We show that autobraking assistance increases human altruism, such as giving way to others, and that communication helps people to make mutual concessions. On the other hand, autosteering assistance completely inhibits the emergence of reciprocity between people in favor of self-interest maximization. The negative social repercussions persist even after the assistance system is deactivated. Furthermore, adding communication capabilities does not relieve this inhibition of reciprocity because people rarely communicate in the presence of autosteering assistance. Our findings suggest that active safety assistance (a form of simple AI support) can alter the dynamics of social coordination between people, including by affecting the trade-off between individual safety and social reciprocity. The difference between autobraking and autosteering assistance appears to relate to whether the assistive technology supports or replaces human agency in social coordination dilemmas. Humans have developed norms of reciprocity to address collective challenges, but such tacit understandings could break down in situations where machine intelligence is involved in human decision-making without having any normative commitments.