I’ve been reading a LOT about the ethics of AI this week. This is something that I am deeply interested in and referring back to my last post, why AI makes me so uncomfortable in theory. The most fascinating of articles has been this one from Mark Ryan (2019). Essentially, Ryan discussed how AI by definition CANNOT be trusted as it “does not possess emotive states or be held responsible for its actions” (Ryan, 2019). This is a fascinating position to me and I think if it is to be believed throws much of the contemporary thinking around AI into question.
In addition, I chuckled when I read the first line of the abstract to this paper on AI and ethics this week: “one of the main difficulties in assessing artificial intelligence (AI) is the tendency for people to anthropomorphize it” (Ryan, 2019). Oops. Anthropomorphizing AI is going to be a large part of my final product.
***Googles “Anthropomorphize”***
Oh, we’re not supposed to?
I’m interested to read and see more about your project!
It is hard not to anthropomorphize something you are having a text conversation with and responding to you. I call it my personal assistant, but it’s not human. It isn’t a moral being, which means we have to be discerning using the information that it provides. In some ways, I feel like it is neutral but I also recognize that the data AI is trained on is not. I think our fear is grounded in how the internet is not neutral and the amount of misinformation on the internet, but the comparison between AI and the internet is not fully accurate. Moreover, you are 1/2 of the conversation, so what you bring to the conversation shapes the responses from AI.