The Female Robot voice: Gender in Robotics and Artificial Intelligence
Alexa and Cortana on our smartphones and computers and Jürgen, Fritzchen and James from our proANT 436 series have one thing in common: They are robot supporters in daily life and are given human attributes through the perspective of their manufacturers and users. The reasons for this can be read in our latest News (robots as colleagues – how a “proANT 436” becomes a “Fritzchen”). This article now deals with the differences between the proANTs and the smartphone assistants, asks why our robots actually get most of the time male names and offers an insight into the topic of gender in robotics and AI.
Stoic transport robots and friendly voice assistants
Jürgen works hard. The transport robot travels non-stop, carrying loads of up to 50kg through VACOM’s production hall, avoiding obstacles and persons, and thanks to an innovative fleet control system, always focuses on its task – a quiet and dedicated worker, apart from occasional warning tones. Alexa is the opposite. The artificial voice has no defined body, her strength lies in her voice, in her kind attention and in her will to help. She listens and is always ready to give a hand to her owner like an invisible secretary.
Both Jürgen and Alexa are in good company. James und Fritzchen also belong to Jürgen’s fleet. Besides Alexa, other assistants are Cortana and Siri. The trend is clear, how we assign our robot assistants male or female features depending on the task. We can define this trend a cliche and criticize it.
Of course exceptions exist. For instance, there is a Siri in male version. However, the tendency is clear, since lots of our customers assign the mostly male names following a complicated procedure with suggestions and employees’ votes.
Stereotypes in robotics and AI
Gender stereotypes come up in the naming. Heavy lifting load robots have a male-dominated job, polite friendly assistants have a typical female job. The forgiveness of names and voices is often adapted to these stereotypes because this role distribution seems familiar. Theoretically, this familiarity could simplify the integration of robots and AI. But actually, this approach is a perpetuation of outdated role models, which can be problematic. Research also shows that gendering robots and AI is complicated and does not necessarily lead to the desired better acceptance. In terms of performance, robots’ gender makes no difference.
The current approach to the gender of robots can therefore be changed, thus avoiding the stereotype problem. This is mainly the responsibility of manufacturers and developers. The lack of diversity in STEM professions is not helpful in this project. This is also noticeable in practice when, for example, speech assistants have difficulty recognizing female voices or face recognition AI only really works for white men.
Solutions for the robot-gender problem
Projects like the genderless artificial voice „Q“ open new possibilities of avoiding stereotypes. More engagement for more diversity in STEM-jobs can also help. A more diverse workforce enable different perspectives to function and design with the result, that test voices and test faces become more diverse. ASTI InSystems support proudly the women in our company with the campaign #ASTIwomen, and the ASTIs STEM TALENT GIRL-program recruits young female talents into the company.
The topic of gender in robotics and AI is surprisingly complex. We transfer our socially rooted gender stereotypes onto robots and AI, this creates problems and may not even be necessary. Diversity in design and development roles in STEM professions is one of the keys to these problems. At ASTI InSystems, we support diversity and welcome applications to ensure that our work is and remains not only innovative but also inclusive.