The results of the study indicate that having an artificial intelligence virtual teammate with a female voice can improve the involvement and efficiency of women in teams where men are the majority. This was discovered through recent research conducted by Cornell University. The study also implies that the gender of the AI voice can have a positive impact on the interaction within gender-imbalanced teams, and may provide insight for the development of AI bots used in collaboration between humans and AI.The results support previous studies indicating that minority team members are more likely to engage when the team includes individuals who are similar to them, according to Angel Hsing-Chi Hwang, a postdoctoral associate in information science and the lead author of the paper.
To gain a better understanding of how AI can assist teams with gender imbalances, Hwang and Andrea Stevenson Won, an associate professor of communication and co-author of the paper, conducted an experiment with approximately 180 men and women who were placed in groups of three and tasked with collaborating virtually on a series of assignments (the study only involved participants who identified as either male or female).
Each group rnrnThere was a setup with either a single woman or man and an additional agent in the form of an abstract shape with a male or female voice. This agent would appear on the screen, read instructions, contribute ideas, and handle timekeeping. However, the bot was not fully automated. In a “Wizard of Oz” experiment in human-computer interaction, Hwang was behind the scenes, inputting lines generated by ChatGPT into the bot.
Following the experiment, Hwang and Won examined the chat logs of team conversations to determine how frequently participants offered ideas or arguments. They also requested participants to reflect on their experience.
“As we observed the actual behaviors of the participants, we began to notice differences in how men and women reacted when there was a female agent or a male agent on the team,” she explained.
“One interesting finding from this study is that the majority of participants did not show a preference for a male- or female-sounding voice,” Won noted. “This suggests that people’s social perceptions of AI can have an impact even when they believe they are not significant.”
When women were in the minority, they were more active in conversations when the AI’s voice was female, while men in the minority were more talkative but were lResearchers found that when working with a bot that sounded male, women were more focused on their tasks. In contrast, men had significantly more positive perceptions of the AI teammate when they were in the minority, according to the researchers. “Simply by having a gendered voice, the AI agent can offer some support to women who are in the minority within a group,” said Hwang.
