Alexa is a virtual personal assistant developed by the company Amazon in November 2014. Alexa performs tasks through voice commands, which provide an integrated and intuitive experience for users. Some of Alexa’s features include: music playback, making to-do lists, setting alarms, and ordering food. The voice of Alexa brings into question the idea of how robots should sound. Alexa’s voice is gendered female, is calm, and formal. This materiality of voice then implies class, race and gender. Do these factors aid people in feeling more comfortable with robots and new technology?
Alexa is a virtual assistant developed to help people by performing various tasks through voice activation. How we engage and interact with it is largely due to its voice . The tone, inflection, gender, and name of Alexa are all elements that construct its materiality. With a different gender or accent, Alexa would no longer be Alexa, thus the script of how people interact with it would be rewritten.
Alexa is a virtual assistant that inhabites Amazon Echo devices. These devices can be described as unassuming matte silver or black cylinder speakers. Their form is ordinary and seems to blend into the environment. Alexa is not Echo; Alexa is more like the intelligence that controls Echo devices. Alexa is the soul, and Echo is the body. This distinction between the physical form that Alexa assumes and its voice become important when considering the materiality of Alexa. This is because the materiality of Alexa has to do with its voice, not the physical way she occupies space. For example, it makes little difference if Amazon’s Echo comes in a spherical shape or a cylinder; however, it does make significant difference if it has a male voice or a female, a Scottish accent or a New York accent, and so on.
When creating any computer intelligence that interacts with people via speech, the voice of the computer is heavily considered. Different tones, pitches, and types of voices make people feel differently and interact differently. Alexa’s voice was created with a combination of stitching together thousands of voice samples with machine learning to make it sound more natural. Alexa has been gendered as female through its voice, her name, and through the way it is marketed to consumers. The combination of all these factors makes speaking with Alexa reminiscent of the feeling one gets when speaking with a motherly figure. Like a mother or caregiving figure, Alexa is always willing to help, it has all the answers, and it is reliable. It can wake you up in the morning, buy groceries for you, play your favorite music, and tell you to grab an umbrella because it’s raining. Users feel as though they are talking to someone familiar with Alexa—an assistant, a friend, a coworker. Our brains have an acute awareness to matching gender and voice. It is rare that when hearing someone speak we are unable to assume their gender. This is also true for Alexa and one of the main reasons why we have yet to see human-like robots without genders.
Alexa’s voice also implies that it’s from a certain class. Due to its proper English and calm, polite rhetoric, Alexa ends up sounding very white and from a higher socioeconomic class. This perception of Alexa would considerably change if users had the option to customize the voice of Alexa. If Amazon had the addition of different voice options with different accents and varying vernacular, the image of Alexa, being a well-to-do proper white female secretary, would be lost.
When considering how many ways you can interact with and customize Alexa to align with your diet, house, shopping needs, and music preferences it feels odd that the option to change Alexa’s voice and name are not an option. It is possible to change Alexa’s name to “Amazon”, so that when waking the device, users would say, “Hey, Amazon…”. However, it still seems odd that Amazon doesn’t allow people to change the name to something a bit more common, such as Alex, for example. For a company who heavily values costumer’s preferences, it might have seemed obvious to include another option other than Alexa who has an equally soothing and calm male voice. Apple’s Siri has the option to be male or female, and because the name Siri doesn’t have a gender associated with it, the transition from female to male voice becomes much more natural and comfortable. One possibility for Alexa’s cemented gender is that Amazon wants users to feel like their talking to a real person. If there was the option of customizing Alexa, some of her humanness might be lost.
The Echo devices that Amazon released are made so that they can be installed into every room in a user’s house. This means that Alexa is able to hear you wherever you are and is listening, ready to help at all times. It is so easy for people to interact with Alexa because there are no new skills users have to learn; users may only use their voice to initiate Alexa’s script. Conversing with Alexa feels so natural because of its familiarity. Just like anyone that you would converse with, the more you do it, the more seamless it becomes. This is the reason why Amazon wants Alexa to be in every room in your house: because the more people interact with her, the more familiar she becomes, the more trusted she becomes.