We were born too late to have attention spans longer than a minute, born too early to see our robot overlords purge us or take us to the stars, but born just in time for ChatGPT to write a solidly 2:1 philosophy essay. With tutorials and a focus on in-person exams, Oxford has probably been less impacted by the boom of artificial intelligence chatbots than other universities. So I don’t want to make some dramatic argument that AI will be its doom.
I’m also not here to accuse you of killing the planet or stealing from the authors whose content was used to train these models, though those concerns are important. My argument is that using AI to write an essay is embarrassing. There’s almost always an element of deception involved, even when you’re not submitting an AI’s output directly. Where use is forgivable, it’s just a bit sad.
Obviously, not all uses of artificial intelligence are the same. At its most involved, it can outright write your essay for you; yet people also use it as a more modest tool or search engine. The paradox of AI use is that almost everyone thinks it’s wrong in some circumstances but can excuse their own use. You might never dream of submitting an entirely ChatGPT essay, but having it rephrase your argument is barely using it. If you do use it for all of an essay, it’s because of exceptional circumstances, and not for an important topic. And if you are regularly turning to it, it’s because you’re at Oxford, and two essays a week is an absurd standard anyway. I’m sure, dear reader, your personal use of AI is entirely unproblematic and, in fact, eminently reasonable. But for everyone else, they should be wary of considering themselves totally free of shame.
Most people, including those doing it, understand the case against using AI content word for word; you’re submitting something that you didn’t produce as your own. As a student who went through however many tests and into however much debt for a degree you presumably (somewhat) care about, you should feel at least a bit embarrassed throwing in the towel. Still, people probably feel the force of this argument less than they would for using essay mills (who must be the real victims of the ChatGPT) – perhaps the burden of writing a prompt makes people feel like they contributed more.
Beyond just copy/pasting AI’s words, you can use artificial intelligence to rephrase your writing, provide a structure or argument, or summarise articles. All these uses involve much more of your own work, and are certainly not as bad as entirely parroting AI. But an essay isn’t just the words on the page, it’s the whole process of research and writing. The argument and response to the literature is the bulk of what matters about an essay, more than the text that contains them. Besides, part of a degree is learning to investigate, to understand jargon and evaluate arguments, to structure cogently and express clearly. If you’re a humanities student, these soft skills are probably most of your degree’s value.
Again, I’m not saying you’re an academic failure if you’ve used artificial intelligence. Oxford is a stressful place, and it’s easy to give yourself excuses. But the freeing part of the Oxford system is that weekly essays just don’t matter that much. If you submit a bad essay, you’ve submitted a bad essay. You might have a rough tute, but nothing will follow you to next week. If you are stressed and time-pressured, you can rush it. It may not be recommended, but it’s feasible to write an essay on the morning of the deadline. Using AI is, to my mind, crumbling under imaginary pressure.
This still leaves some uses of AI that aren’t really deceptive, like using it to explain a concept or test an idea. These are the sorts of uses I imagine the University means by AI being a ‘supportive tool’. Using AI like this is not necessarily wrong, it’s not really more deceptive than talking to a friend about the topic. But why not talk to an actual friend? Oxford is an academic community full of people who love to talk, many of whom are thinking about the same things as you, or have been recently. Even people doing other subjects can provide insight. Why replace human interaction with a chatbot gobbling resources to do its best impression? The unique value of AI is preventing you from leaving your room.
Everyone draws the line of acceptability just below what they do. You can always imagine someone doing something worse, someone relying on AI more, or more often, or for an easier degree at another university. The fact is that if AI is doing any work, you’re passing that work off. If it’s not, what’s the point?
Have an opinion on the points raised in this article? Send us a 150-word letter at [email protected] and see your response in our next print or online.
