When my classes have, over the years, touched on the beginnings of the conquest of the Americas, I am always struck by the incommensurability of the experience. It is hard to even begin to find the ways the Spanish and Aztecs lacked understanding of each other. The Aztecs would never see Europe or learn about its history or customs; it might as well have been the realm of the gods for all the reality it held for them. Similarly, the Spanish were confronting another world of people, empires built on thousands of years of history they could not understand. Different ways of looking at the world collided on the levels of understanding time and forming social meaning. As a way of understanding what this initial phase of confronting the Other might have felt like, let us consider the next time it is likely to happen -- artificial intelligence.
By artificial intelligence I don't mean the Hollywood robots that fall in love or rewrite Shakespeare, but rather the ever-advancing algorithms that are being pursued at a breakneck pace around the world. One example I would use is Alexa (Amazon's robot assistant). It was recently revealed that Amazon has 10,000 employees working on Alexa, a massive number that has doubled in the last year, and the company's CTO talks on his blog about his desire to create more "human" interactions with Alexa, with a near-term goal of the program being: "a socialbot that can engage coherently for 20 minutes in a fun, high-quality conversation on popular topics such as entertainment, sports, technology, and fashion. " (source). At first this seems like a fun gimmick, as do proposals like Google's to have its AI assistant navigate booking appointments. Yet the concept of a program replacing a human in social interaction quickly becomes troubling. What if Google's assistant learns that being mean to restaurant staff is more likely to yield a reservation, or Alexa learns to talk politics by telling people what they want to hear? Yet these are only the most immanent scenarios, and the real comparison to Cortes and the Aztecs comes when these programs can converse indefinitely and without constraints. As we have seen when AI reaches the top levels of human skill in chess, poker, go, or even video games, more than simply replicating human tactics, it creates its own. What will that look like in the field of language and communication? AI might become inhumanely good at reconstructing our psychological profiles from minutes of conversation, finding new logical fallacies or biases within the human brain. Talking to a sophisticated AI might become as asymmetrical as the Aztecs, who always tell the truth, talking to Cortes and having no concept of falsehoods. Are these two experiences meaningfully similar? Perhaps not, but the comparison illuminates the emotional reactions we might have when speaking to something completely outside our usual systems of understanding. All of this begs the question of how people should respond to such circumstances, Particularly in a context such as the one the Aztecs faced where confronting the Other was not a choice, but an inevitability. Would we expect ourselves to be able to access new forms of understanding in order to confront an adversarial AI? The thought gives me more sympathy for the Aztecs, as such a response would require highly abnormal self-determination in an individualist society, let alone a more collectivist one. Certainly, the challenges Todorov points to in the book do not seem easy to overcome, particularly when we imagine ourselves in a similar situation.
0 Comments
Leave a Reply. |
AuthorAmerican University Class of 2021. Interested in state-building and economic governance. Archives
December 2018
Categories |