WICKEDLY CLEVER WIRINGS* : GRADUATING CLASS A.I. ESSAY

THE TECHNOLOGY behind A.I. is wickedly clever, and back when Large Language Models were still a research experiment - before they propped up the U.S. economy.

My friends and I remember trying to explain to my grandfather, who has since died, that '' backpropagation, '' a technique vital to A.I. grew out of attempts to quantitatively prove Freud's theories about the '' flow of psychic energy.''

I don't think I really sold Grandpa on why he should care - but to me, the development of A.I. was human genius at its finest, and I couldn't wait to open arXIV links people would text me containing the latest and greatest research.

The output of a model didn't matter anywhere near as much as how it was designed.

Now, the opposite is true. A.I. is an application that people actually rely on, and companies have become less and less transparent about its design.

What counts is the immediate response you receive when you send a reading to ChatGPT on your walk to the class.

Most students call OpenAI's model '' Chat.'' Many refer to it familiarly, consulting with Chat repeatedly over the course of the day, letting it decide how to text a situationship and confidently repeating hallucinated assertions while in line at the coffeeshop.

For years, online livestreamers have used the word. '' Chat '' to interact with their audiences, asking commenters to tell them what choices to make in video games.

That students now use the same name for A.I. feels appropriate. What really is the distinction between a nameless, faceless human you'll never meet except over the internet and a statistical approximation of the same thing?

The internet has already allowed us to feel more connected than ever. A.I. lets us cut out the human part of human interaction entirely?

This Master Essay Publishing continues. !WOW! thanks The New York Times and Theo Baker.