An Ode to ChatGPT: Please Don’t Replace Me

Good Writing will Always be Human Writing

M. D. Kwak

Supposedly, this thing’s going to be replacing our jobs soon (oh the joy of being discarded as a worker ant in the capitalist machine!). I can’t decide whether to be more depressed about my diminishing employment prospects or by the utter pointlessness of most human occupations.

It’s especially horrifying when this glorified jumble of code can’t spit out a half-decent discursive piece without sounding like an overzealous fourth-grader having discovered their newfound capacity to use the internet. That’s why I scoff at the eulogies of tech writers who somberly predict the obsolescence of ‘human beings’ at the hands of AI. If it can’t help me with English homework, it might as well be useless.  

Maybe a day will come when we outsource writing to AI like how we outsourced Maths to calculators or transferred our collective knowledge to the internet. I guess that’s also the day I put down my pen for the Record (oh – the horror!) and let a glorified text prediction machine do all the work.

Perhaps my sarcasm (no doubt inferior to ChatGPT’s excellent humour) illustrates my disdain for the hype train surrounding AI chatbots. GPT can’t even exist without us anyways. The way most AI chatbots work is by feeding off an existing corpus of data; all its algorithm does is decide what word will come next in a sequence. For now, chatbots can’t actually learn anything new; it is dependent on human input. That means scientists and researchers still need to discover more about our universe and human nature. Authors and essay writers still need to create and analyse literature. Of course, this is all so that their work can be unceremoniously cannibalised into a GPT response (what’s intellectual property, am I right?).

But I can’t deny that these chatbots will only get better and better. Unfortunately, I can’t say the same for human progress. Not when TikTok has (quite impressively) curtailed our attention spans to near zero and limited our empathetic and communicative capacities to memes that defy all conventions of humour or political correctness.

I wonder if we’re a bit sanctimonious of our own ability to create truly original things. Because there’s nothing special about our capacity to simply observe and write down our thoughts, feelings or scientific conclusions based on sensory inputs. GPT-4 can now describe an uploaded image. Machine learning can use sensors and recognition code to mimic our visual prowess. If AI had access to the world’s security cameras or the personal data of millions of people, it could probably be more functional than any Buzzfeed writer or sociology researcher.

What’s left I guess, are expressions of feeling, style, and the human mysteries of the creative process. Of course, that’s not to say AI can’t replicate our cognitive or psychological processes. What’s to say emotions can’t be grown out of a lab or birthed from pages of code?

I’ve come to accept that I function like an inefficient version of GPT. My sentences are not generated by AI, but they are largely the synthesis of my favourite authors. My “personal style” is a bastardised and inferior amalgamation of Orwell’s essays, Shakespeare’s poetry, and of the phrases my favourite teachers and friends use. Perhaps there is some spiritual self in all this, but I cannot locate it. Our ‘code’ is the complex string of DNA; our prompt is the entirety of the world.

Why then do I disdain the writing of AI? Why am I lured again and again into attempting to decipher Hegel’s densely written philosophy or proclaim (to no one in particular) my attempts to read Dostoevsky for the fifth time?

I think I enjoy reading human writing because I like getting mad at people (as fun as hurling abuse at a chatbot). Perhaps the entirety of human creativity could be replaced with an immense surveillance state and a GPT-4 plug-in. But the reason we listen to songs and read books and look at paintings is to see the self in another self or even to just see what other people are capable of creating. We feel strangely comforted by the knowledge there is someone behind that artwork and in many ways their story becomes part of the human story and therefore our own.

The short answer: human writing is good because it is human.

But crucially, I believe human writing exists for the author as much as it does for the reader. It is their vassal for self-expression; a searching cry that seeks validation for one’s existence, a confirmation of one’s autonomy. For many, it is a coping mechanism for when the horrors of life get too much. And for me at least, it’s the most steadfast way I can strive for self-improvement. Because more than reading, writing is thinking and thinking is what makes life interesting.   

So, when GPT comes along and writes for the Shore Weekly Record instead of me, the loss won’t be to you, the readers. After all, you’ll be reading quite similar works. The loss will be felt by me. No longer will I feel the cringe and joy of penning a transcript of my rambling thoughts. No longer will my brain strain under the pressure of deadlines and the persistence of writer’s block. And no longer will I endeavour to extend my knowledge, to think critically in ways I otherwise wouldn’t have, about things that would have slipped past my vision.

So, I beseech you AI. Please don’t replace us.