Am I AI?

Over the course of the year, a couple of things have stood out to me in regards to large language models and how they influence the way we operate. I’ll begin by describing where it affects me most and spiral off into some more subtle notes about my observations.

The word maze

As part of my role, I have to write a lot, be it design docs, technical articles, or describing my thought process about a given problem, and in more than a few instances, i have caught myself thinking. “Hope I don’t sound like an LLM?” Not because my response was pasted from ChatGeePeetee , but because i tend to write how i speak and sometimes that is in a monotone voice, now thankfully a lot of the people i interact with frequently know how i speak so it’s not shocking to them, but i can’t help but wonder when i started to rethink the tone of my responses.

Unseen influences

When “Experts” or enthusiasts say Large language models and AI (they are mostly refering to LLMS tbh), are going to change the way we think or interact with one another i can’t help but wonder if they realize its likely in the way they are often describing.

I read a really good blog the other day about reading to forget, aside being how I operate, a lot can be said about how the text you read influences your speech and thought patterns. Take slang/internet abbreviations, at first you might not know what “lol” or “lmao” means, but someday you wake up and it feels like you have used them forever.

As the internet fills up with more AI slop, I expect some of the models vocabulary to invade our speech, however, how that will turn out is a topic for another article. While writing this i came across an article from The Verge which seems to support some of my thinking here.

Standing out

If we all sound the same and all my work is going to be training data, I should try and stand out, right ? Well, for me, I don’t know, I have always maintained that English and by extension language is a tool for communication, I like to pick and choose how I speak at any given time, this includes the words and the era they are from. This has landed me in some debates at work as English is not everyone’s first language, which is important to remember when writing for a global audience.

It is within this mashup of words that my personal voice lies, whether or not Mr ALTernative to man is able to replicate this or not,only time will tell.

As I come to the end of this rant, I do want to note there is some genuine slop of literature out there, and it sucks to read, but there aren’t many good solutions either. Your “humanizer” and “Checkers” aren’t a solution but rather a symptom of a bigger problem.

And so i ask once more, Am I AI? Was this piece an LLMs latest test run?