Technology should adapt, let us pick who we hear
It may be easy to draw the conclusion that the voices of these AIs are a reflection of the stereotype that most assistants are women.”Sam Ramirez, El Leñador writer
In the past few years artificial intelligence (AI) has become a topic of intrigue as many companies have replaced traditional algorithms with AI, machine learning, neural networks and other methods to achieve a computer-generated program.
Yet most people aren’t aware of these AI, and the ones we are exposed to usually come in the form of an assistant. Examples of these assistants are Apple’s Siri, Amazon’s Alexa, Microsoft’s Cortana, and Google’s Google Assistant.
While they help make our lives a bit more organized, there is an overlapping trait with these AI assistants, they come with a common setting, a synthesized female voice.
It may be easy to draw the conclusion that the voices of these AIs are a reflection of the stereotype that most assistants are women.
But in a journal published in 2011, Computers in Human Behavior, a study by the Indiana University School of Informatics set out to see the effect that an AIs voice has on people. Specifically they tested whether men and women prefer a female or male voice in virtual voices. What they found is that both men and women preferred the female voice, which is supported by an older study conducted in 2004, published in the Journal of Personality and Social Psychology and that bias translates over to virtual voices.
But this doesn’t apply as a general rule, as IBM’s Watson AI, who competed against and beat a champion Jeopardy winner, has a male voice and is not an assistant. So there are certain scenarios where the bias is made apparent.
Overall, people will apply pronouns to any machine.
The default voices for many of these AI assistants are a female voice and the
reason could be anything and may be reinforced by the stereotype that people find a female voice more nurturing and calming.
But the fact remains that in the current social and political climate we should have the option to pick what voice we would rather hear.