
Although it is justified to dismiss artificial intelligence and large language models as exalted extensions of auto-correct and predictive text—autocomplete—there is a danger is dismissing the analogy that a chatbot is a mere
calculator of words. Albeit an adding machine has
unimpeachable and unbiased output, AI too has by studying frequency a handle rather than an understanding of custom through patterns that even linguists have not been able to precisely
pin down, which despite no understanding can through brute force pass the Turing test and even on a
rudimentary dataset become convincingly fluent, just enough so as the technology and expectations advance.