This paper explores the workshop as a tool to raise awareness about unconscious gender bias in AI voice assistants. To answer the research question of how tools can develop a norm-critical understanding of gender biases and stereotypes in AI voice assistants, a workshop was conducted. The workshop was executed with Computer Science students focusing on their perception towards voice assistants, different gendered voices, including a genderless voice, and the connection between biases and stereotypes within norms created by society. The results showed that the characteristics that students associated with female and male voices reflect earlier studies and are consistent with societal gender standards. Despite having people from many cultural origins, this demonstrates how deeply the common concept of gender norms are rooted. Participants were forced to confront their own underlying prejudiced thinking by pushing them to play extreme stereotypes, ask critical questions, and participate in a collaborative conversation. During the conversation, the students began to consider the origins of gender norms in society. After hearing a female, male, and genderless voice, all of the participants favored the female one, indicating that the genderless voice still need tweaking and improvement to meet market demands. Essentially, using a norm critical approach, the workshop forced unconscious biases to become conscious, spoken about, and collaboratively discussed. The students left the workshop feeling more aware and resolved to reflect on their unconscious gender thinking, according to the results of the reflective survey.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:liu-186187 |
Date | January 2022 |
Creators | Schumacher, Clara |
Publisher | Linköpings universitet, Institutionen för datavetenskap |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0016 seconds