The Line Between Human Language and ChatGPT is Fussier Than You Think
When asked to check the box to confirm that I am not a bot, I do not give it much thought. Of course, I am not a robotic. When my email client suggests words or phrases to finish my sentences, or my phone guesses what I’m going to text next, I begin to doubt myself. Was that what I intended to say? If the application had not suggested it, would I have thought of it? Am I a robot? These large language models were trained using massive amounts of \”natural human\” language. Do robots have a human component?
AI chatbots may be new, but public discussions about language change aren’t. As a linguistic-anthropologist I find the human reactions to ChatGPT most fascinating. These reactions reveal the underlying beliefs about language that underlie people’s ambivalent and uneasy relationship with AI interlocutors.
ChatGPT, and similar tools, reflect the human language. Language is both a highly original tool and a very unoriginal one. Chatbots reveal the same tendencies and patterns as humans.
Source:
The Boundary Between Human Language and ChatGPT Is Fuzzier Than You Think
