The Social Uncanny Valley

The Uncanny Valley effect refers to the observation that robots who are highly human-like elicit negative responses from observers. Based on this observation, we examine human responses to AI-generated objects which are very similar to those made by humans. To date, it is unknown whether such products can elicit the Uncanny Valley effect. Focusing on AI-generated art, this research explored the relationship between AI products, the Uncanny Valley effect, and people's perceived social competition toward AI. Study 1 generally supported the "social uncanny valley effect." We found that AI-generated products aroused the Uncanny Valley effect under certain circumstances, and people had more sense of competition toward AI when the human-likeness of those AI-generated products increased. Study 2 remedied some of the limitations of Study 1, and further supported our hypotheses. This means that people will have a stronger sense of creepiness toward AI and a more intense sense of competition against AI because of the increasing human-like factors of AI-generated products. Our research suggested that people should re-think the current AI developing trend to avoid a new style of social conflict between humans and AI. This conflict might be alleviated by emphasizing the assistive nature of those AI rather than replacing humans with those human-likeness.

Level thesis:degree_level

Master's Degree

Year dc:date.issued

2023

Author dc:creator

• Wu, Di

Advisor dc:contributor.advisor

• Kemmelmeier, Markus

Committee members dc:contributor.committeemember

• Bok, Jared
• Bogard, Rebekah

• Art
• Artificial intelligence
• Uncanny Valley

Handle dc:identifier.uri

http://hdl.handle.net/11714/10543

OAI identifier oai:identifier

oai:scholarwolf.unr.edu:11714/10543