AI is becoming more and more prominent in everyday life. There are rumors of people using it to cheat on homework and have essays written for them, so I decided to test just how effective AI really is. 

AI can be used for CompSci homework because it is good at calculations and coding, but is it able to effectively complete an assignment based on human emotions and morals? In English class, I have to submit pieces of writing called OPs. They are graded more on having a moral than grammar and structure. 

I wanted to try using Chat GPT, a new popular AI website that can create essays for users. It is being promoted on popular social media sites like TikTok and has probably been used numerous times in place of student work.

I entered a prompt on ChatGPT about the importance of honesty. I then changed the prompt to be about the importance of honesty after cheating on an essay. However, this didn’t alter the results at all, so I was left with the same essay with an assortment of facts on the importance of honesty.

The essay seemed like a lot of fluff since the AI didn’t have real-life experiences on which to base the lesson on. It also didn’t have a clear conclusion because it continuously wrote until I clicked stop. Despite my disbelief that this could pass as my own writing, I decided to submit the AI-written work. 

I approached my English teacher and asked him to read the essay  and  to tell me what I would receive on it if I turned it in to take the place of an essay I had written prior. 

He essentially told me I would receive a 20/20 because that’s what the assignment was worth and it wasn’t graded strictly. 

Due to the lenient grading, this experiment didn’t really show that AI can write something with as much emotion as a human can. However, it did exemplify that something that took me about an hour to write could have been done in less than one minute and receive the same grade. 

Later on, I did end up telling my teacher that I had shown him AI work and he was mind-boggled. He noted that the work I had shown him would’ve received an A grade but it wasn’t as heartfelt or entertaining as the student work that was done with effort and based on real-life experiences. 

Although this article made ChatGPT sound like an easy and effective way to cheat on an assignment, teachers are already starting to catch on. Students are always one step ahead of teachers, and every time they find a new way to cheat, the teachers find a new way to prevent it. In this case, educators are now strategizing to detect the use of AI in the place of student work. For example, there is a new website called GPTZero created by a 22-year-old Princeton student. It was designed to detect work that was written by AI.

I submitted the ChatGPT piece that I had turned in to my teacher, and GPTZero detected that it was written entirely by AI. 

Then, to test if this website was truly accurate,I also submitted a piece of writing that I wrote myself. It came back as entirely written by a human.

This goes to show that although AI seems like an easy way to cheat, it is still very risky.

DISCLAIMER: This article does not condone cheating with AI.