One day, the devices began to talk and address the word to the people. The Bluetooth box, for example, which always announces the status of the battery level in a robotic female voice immediately after it is switched on, and then, after sometimes a longer pause, announces which devices in the immediate vicinity it has just connected to . At the beginning of the relationship with this jukebox, the artificial voice primarily caused astonishment. Siri, Alexa, Cortana and all the other language assistants from the big tech companies sound a lot more human. A phenomenon that has become part of everyday life: the chattering devices almost sound like real people. Recently, an engagingly friendly male voice guided you through the menu items on the touchscreen of a photo booth in a suburban shopping mall. It worked so smoothly that a chummy proximity to the machinist from the digital abdomen of the passport photo booth set in.
This is exactly a topic that scientists are also concerned with. When the devices talk to people, the basic question is how do those involved in this conversation understand each other and under what conditions does a positive exchange take place? Weizi Liu and Mike Yao from the University of Illinois Urbana-Champaign now report in the journal Computers in Human Behaviorthat the gender of everyone involved is an important factor. Women and men react differently to speaking devices – and above all it matters whether the artificial voice sounds like a woman or a man.
Women generally rated the exchange with a chatbot more skeptically and less positively than men – regardless of how it was configured. However, they gave the worst rating when the device spoke to them in a female voice. On average, they found a device that sounds like a woman to be the least likeable and well-disposed. Men rated the chatter devices significantly more positively across the board, regardless of whether the voice sounded neutral, female or male. However, chatbots that spoke in male voices aroused the highest sympathy values among them. In this constellation, the chemistry was just right, according to Liu and Yao. One study participant was so enthusiastic about a male chatbot that the scientists had to ask him several times to finally end the conversation because the next test person was waiting.
Nevertheless, most of the subjects stated that they would prefer to configure Alexa, Siri or other speech boxes with a female voice if they had the choice. This attitude corresponds to what has long been considered textbook knowledge in psychology: Announcements, hotlines, chatbots and other devices should speak in a female voice, because this would be rated as warm-hearted, trustworthy and amiable. Men’s voices, on the other hand, arouse associations such as “aggressive” or “dominant” and should therefore be avoided, it was said for a long time. So isn’t that right? Who knows. One thing is clear: in the past, when the devices only beeped gender-neutrally, the world and the relationship to the devices were a bit clearer – that was just annoying.