Loose Notes, Serious Questions

My professor marked my essay using AI – help!

Bright young boy in front of a blackboard filled with formulae, representing intelligence, education and AI-assisted learning.
Photo by Vitaly Gariev on Unsplash
Key takeaway: If students submit essays written with AI and lecturers respond by marking them with AI, university assessment risks becoming a polite exchange between two machines. The real issue is not only cheating or efficiency, but whether students and teachers are still doing the intellectual work themselves.

 

Introduction

You know, the only thing more depressing than an essay written using AI is when the lecturer marking it is also using AI.

To be fair, let’s try to understand this from the professor’s perspective.

Academia, pre-AI

To get to where they are today, they would have had to write a PhD thesis of around 80,000-100,000 words, long before AI was invented. Many of them did this the hard way – without the assistance of ChatGPT, Google, or even the Internet.

Basically, they wrote their thesis on a battered laptop at a non-ergonomic desk, surrounded by unruly piles of hard-copy books and journal articles. This is exactly the same way they drafted every single undergraduate essay before.

After their thesis was written, they spent several more years trying to turn it into a book, again without the help of AI. Without a steady income, many of them would not have even been able to afford a copyeditor, even at half price.

If, by some miracle, they were lucky enough to secure a tenure-track position, they then expended a great deal of blood, sweat and tears to obtain the holy grail of academia: tenure.

Many happy years of research, conferences, publications, and book launches would have followed, punctuated – for better or worse – by a few hours of teaching each semester.

The rise of AI slop

Then, suddenly, a new generation of students emerges, one that seems no longer to rely on its own brains and instead submits essays that are uniformly composed, balanced, and fluent. No spelling or grammar mistakes, no infelicities of expression, and not even a single missing or redundant apostrophe. Each sentence is confident and well written, and it follows the previous sentence in a more-or-less orderly fashion.

Of course, the tell-tale signs of generative AI slop are obvious to all.

For instance, there is often no sustained persuasive argument. In a traditional essay, an argument is carefully developed over the course of the essay, culminating in a strong, even if naïve or flawed, conclusion.

Another giveaway: liberal use of em dashes.

The most blatant giveaway of all? A stray line inadvertently left at the top of the essay, complete with an em dash and awkward use of bold text: “Sure—here is a 3,000 word essay answered like a University student:

Fighting fire with fire

Confronted with AI slop, can one really blame one’s professor for throwing their graduation cap on the ground in disgust, cranking up their ChatGPT and fighting fire with fire?

In this scenario, the whole process will be little more than a polite intellectual exchange between two AI systems, well above the pay grades of both student and professor alike.

Scroll to Top