Social norm learning from non-human agents can induce a persistent perceptual bias: A diffusion model approach

Area of Science:

• Social Psychology
• Human-Computer Interaction
• Cognitive Science

Background:

• Classic social influence studies focused on face-to-face human interactions.
• Computer-mediated communication and AI agents are increasingly prevalent.
• Understanding influence from non-human agents is crucial in modern society.

Purpose of the Study:

• To determine if non-human agents can exert persistent social influence.
• To investigate if non-human agents contribute to the emergence of social norms.
• To explore the cognitive mechanisms underlying non-human agent social influence.

Main Methods:

• An online adaptation of a social norm learning paradigm was employed.
• Diffusion model analysis was used to examine cognitive underpinnings.
• Participants interacted with non-human agents in a controlled online setting.

Main Results:

• Non-human agents demonstrated the ability to induce lasting social influence beyond immediate groups.
• Evidence suggests non-human agents can contribute to the formation of social norms.
• Diffusion model analysis indicated shared cognitive mechanisms for human and non-human social influence.

Conclusions:

• Non-human agents are capable of establishing persistent social influence and contributing to social norm emergence.
• The cognitive processes engaged by non-human agent influence are largely consistent with those of human influence.
• This research highlights the significant impact of AI and digital agents on social dynamics.