Part of the overall argument of my dissertation is that the true threat of automated technology, AI, robots, etc., is not necessarily that robots are becoming dangerously human-like and that we might soon face a Matrix-style machine revolution, but rather that by increasing our interaction with automated, robotic technology we become more robotic.
I have never felt this sensation more acutely than when trying to call any kind of customer service line these days. It’s almost impossible to talk to an actual human person without first wading through a byzantine series of automated voiceover machine prompts. “Please describe your problem in a few words,” demands an affectless voice. I then proceed to ramble incoherently and in a totally unnatural cadence that results in the machine saying “I’m sorry, I don’t understand.” It’s maddening not only because you’re not solving your problem but also because you don’t feel or even sound like yourself. Seriously: next time you’re interacting with a robot voiceover, pay attention to the way you speak. It sounds more like you’re typing words into a search bar than anything resembling human dialogue; it is devoid of grammatical connector words, just a series of keywords rambled off in a frustrated tone. Never mind that the reason for the call in the first place is not easily summarized in simple sentences. Interestingly, a problem I can’t solve by Googling and that requires a phone call somehow never seems to fit into the simplistic categories that the robotic voice asks me to fit it into. Quelle surprise. But understanding poor descriptions of complex problems that defy easy categorization is something humans are very good at!
Speaking with an automated phone directory literally brings us down to the robot’s discursive level and forces us to talk and communicate like a robot does, because as humans our default social mode is to maintain interlocutor equilibrium. We subconsciously find and convert ourselves to the least common denominator in a communicative exchange. But robots can’t change their communicative register, so the only way to achieve mutual intelligibility with a robot is if we ourselves speak robotically. I find the whole experience of automated customer service incredibly frustrating and illustrative of the real threat automation poses, at least in the short term: It’s not that the robots will become like us, it’s that we will become like the robots. And of course, once we have normalized robotic human behavior, it will be much easier for differences between man and machine to cease to exist.