Social agents, stereotypes and adoption : Exploring the effects of stereotypes on social agent interaction and examining patterns of social agent adoption

Detta är en Kandidat-uppsats från KTH/Skolan för elektroteknik och datavetenskap (EECS)

Författare: Hugo E. Norberg; Karim Nettelbladt; Philip Nilsson; [2023]

Nyckelord: ;

Sammanfattning: The rapid advancement of artificial intelligence and robotics has opened up new possibilities for the creation of human-like robots and social virtual agents. As these lifelike robots continue to be developed, it becomes crucial to investigate the potential impact of stereotypes on the interactions between humans and robots. This study aims to explore two main objectives: to find out if a virtual agent's appearance and ethnic representation influence interaction and to explore patterns of adoption of social agents. In the first study, 32 participants interacted with a virtual agent using a videoconferencing platform. The participant was assigned one out of two virtual agent conditions, one with a typical Swedish face and accent and another with a typical Kenyan face and accent. The results showed that the Kenyan agent was rated as having higher English proficiency than the Swedish agent. On the other hand, the Swedish agent was perceived as more authentic than its Kenyan counterpart. In the second study, a short questionnaire was distributed to a sample of 100 participants to investigate their utilization and frequency of engagement with social agents. The results show that adoption of social agents for work or study have entered the “Late Majority” stage. Furthermore, an inverse relationship between age and the adoption of social agents was identified.

  HÄR KAN DU HÄMTA UPPSATSEN I FULLTEXT. (följ länken till nästa sida)