We use virtual assistants (VAs) because we’re lazy. Well, no, that’s not entirely true; but we employ our VAs with all the tasks we can’t (or don’t want to) be bothered doing ourselves. In theory, dictating tasks to a computer that is always at your beck and call, never gets annoyed with your constant demands, never gets tired, never talks back, and seemingly knows everything, is a pretty good deal. However, for anyone that’s ever used a VA, you quickly realize they aren’t as foolproof as advertised.
Top corporations, such as Amazon, Google, and Apple, have all come out with their own version of the VA; Alexa, Echo and Siri, respectively. The entire appeal of VAs are their offered transparency. Going all the way back to the first week of our Digital Studies 101 course, we learned that functioning at its optimum, technology is ideally transparent; it should fade into the background. One very clear aspect of technological transparency is its ability to be consumer friendly. By that I mean it should simply be easy to use, and using it should be an automatic process. The moment VAs are unable to recognize, understand, interpret, or fulfill our requests, their transparency is shattered. We are forced to repeat the request 2, 3, or even more times; that totally undermines the rational of using VAs: if I wanted to do that much work, I would have just made the reminder, for example, myself, or I would have asked someone else to do it for me. At that point, VAs are the opposite of transparent, their inadequacies and shortcomings are glaring obvious. What’s comical however, is that I’ve personally witnessed people repeatedly speaking to VAs, hoping that they’ll get it right eventually, instead of just doing the task themselves. My mom has repeated a message she wants Siri to transcribe and send 10 times, I’m sure she would have continued if Siri hadn’t gotten it right on the 10th try. She would rather do extra work by repeatedly saying the same message than just type the message herself.
Continuing on the topic of breaking the offered transparency of VAs, I recently had an experience with my phone that made me want to throw it at a wall. I wanted Siri to delete some memos that had been piling up. I asked her to delete all my memos, she said “there are no memos named ‘all’”. This was beyond frustrating because any other human, they would know exactly what I mean by “delete all my memos”, there’s a commonality of language that humans have that computer-human relationships lack. For anyone that has ever coded, you know that computers follow a very strict and logical way of “thinking,” they do not truly understand linguistic expressions, when they ask you which memo you want to delete, they are waiting for the name of a memo, nothing more.
Until VAs master the ability to truly understand rather than just simply recognizing and fulfilling tasks, this will be a reoccurring issue. VAs that can understand and become “conversational computers” (Finn 2017) bring about a whole other set of issues and concerns associated with autonomous technologies, we’ve seen enough movies that clearly depict the acquisition of self-awareness by computers as the beginning to the end of the human race.
I guess we’ll just have to wait until technological singularity to find out.