In recent months, you may have heard of a program called ChatGPT. Released to the public on November 30, 2022 by a company called OpenAI, the bot has seen a massive surge in users. It’s very easy to use: just give it a prompt and it spits out a response for free. While this is a very interesting and potentially useful tool, it has also raised significant concerns about students using programs like this to cheat their way through assignments.

I tried out ChatGPT for myself to see how successful it could be with a prompt from my AP English class (don’t worry, I had already done the assignment on my own). The program is blocked on school Chromebooks, but I could easily bypass this by running the program on my phone and emailing the response to my Chromebook. In less than one minute, the AI spit out an essay that can be read here. It’s a bit short, but overall it’s a pretty great essay. In some ways, it may be even better than the one I wrote.

I then ran this essay through multiple plagiarism checkers. Small SEO tools marked the essay as 100% unique. DupliChecker said the same. When I tried out GPT Zero, a program designed to detect AI writing, the essay was thankfully marked as “most likely AI generated.” However, with just a few tweaks to the essay, I was able to get it cleared as “most likely written entirely by a human.” While programs like this are incredibly useful, they are not as infallible as people may think.

What does this mean? Well, first of all, it’s a little scary. There’s something very disconcerting to me about a robot that can write something so thorough in such a short amount of time. This is also very concerning in terms of cheating. Students now have the power to bypass assignments using AI, which is much harder to detect than traditional methods of cheating. While it’s definitely possible to guess that something was written by an AI based on the formulaic sentence structure and content, it’s harder to prove that this is the case. Even with programs like GPT Zero, things can still slip through the cracks, especially if the teacher grading is unfamiliar with AI and programs that can detect it.

So what’s the big idea? Because right now, this article just seems like a “how to cheat on your essays” manual. My ultimate goal with this article is to make teachers and anyone else who has to read things like essays aware of this new tool and ways to spot it. Do I think AI is going to destroy the school ecosystem and take all our jobs away? No. Or at least not yet. However, it is something to be aware of. There are ways to detect AI generated text, but they’re useless unless people know about them. 

And finally, if you are considering using ChatGPT to attempt to cheat, I want to urge you not to. Not only is it morally wrong, in my opinion at least, it’s not very smart. Teachers know what your work looks like, and even if they aren’t familiar with AI, they’ll be able to tell that something is up. At the end of the day, you’ll just make a fool of yourself when you’re caught. Trust me, just this year there have been cheating incidents in two of my classes and both times they were caught and everyone in the class faced consequences (writing essays by hand rather than online, having to re-do full assignments, etc). 

There are conversations to be had about how programs like ChatGPT can revolutionize the ways in which we work and communicate, but they will go nowhere if people are only interested in cheating. If any teachers were to read this article, I would encourage them to be aware that this is a possibility.

Just for fun, I had ChatGPT write its own version of this article, which can be read here. I think mine is a bit more thorough and has a more personal touch, but I suppose that’s up to the reader to decide.