The new AI ChatGPT has the potential to change the Riverside English department.
The free website, launched in November 2022 by OpenAI, can write an essay nearly instantly with a simple prompt. Having scanned the entire content of the internet from 2020 through 2021, ChatGPT can write about virtually anything.
“Talk to the computer… and get what you want,” said OpenAI CEO Sam Altman in a November 30 tweet.
For example, ChatGPT can create a song in the style of Taylor Swift, an essay about the symbolism of the novel Lord of the Flies, and even, according to English teacher Matt Smith, “a guide on how to get a sandwich out of a VCR written in the style of the King James Bible.”
Teachers and students have mixed feelings about this new technology.
“I find it fascinating that machine learning has gotten so good so fast, that some of it is legitimately good,” said Smith. “It really intimidates and scares me, because if I’m trying to get humans to do writing, and they have the easier option of getting a machine to do it, I don’t know yet how I’m gonna navigate that.”
“It is a very cool breakthrough in technology,” said senior Wade Gabriel. “But I feel like its accessibility is also a hindrance in some ways because it’s so easy to cheat and to use it for malevolent purposes.”
“While it’s advancing artificial intelligence, it is doing so at the cost of making people think critically and share their own thoughts,” said English teacher Emily Ericson.
Some Riverside teachers expressed concern that students would use ChatGPT to cheat.
English teacher Laura Brady worried she would not be able to detect an AI-written essay.
“I imagine that an effective AI could write something that sounds authentic to someone’s voice and captures an idea that does not exist easily on the internet,” Brady said. “It makes the prospect of assigning online work very daunting.”
However, other teachers were confident that they would know if a student plagiarized their essay with ChatGPT.
“I’m always reading what [my students] are writing during in-class writing assignments, and I know their writing voices and styles,” said Ericson.
“I would be able to tell, if I had one sample of a student’s writing, if their follow-up work was genuine,” said Smith.
However, Smith worried that if the first essay a student turned in was AI-written, then he would have no way of determining their writing style, and might mistake future essays as plagiarized.
Many teachers and students think that the quality of essay writing from ChatGPT is not up to par with the abilities of students.
“[ChatGPT] has been shown to have continuity and be able to produce a basic rundown of a topic, but it can’t mimic writing style,” said Gabriel.
“Different teachers have shared with me examples of AI writing, and it’s not all nonsense, but it’s also not all coherent,” said English teacher Victoria Watson.
Ultimately, students and teachers agreed that an AI-generated essay might not have the level of creativity, critical thinking, and nuance that a human can produce.
“Part of the reason that English and Social Studies are part of the humanities is I think there’s something essentially human about language and speech,” said Smith. “I’m worried that when you remove that human element, you’re losing an essence of the discipline.”
Can an English teacher detect an AI-written essay?
Artificial Intelligence doesn’t scare English teacher Mira Prater.
“I can tell when a students’ writing is not their own,” said Prater. “It’s very clear when it’s not a student’s own voice.”
The Pirates’ Hook decided to put this to the test. With Prater’s permission, the Hook planted an AI-written essay in her class.
On February 10, Hook reporter Jackie Larios Dominguez, who is also a student in Prater’s Honors English III class, turned in an essay about The Great Gatsby created by ChatGPT.
Prater detected it immediately.
“I thought, as I was reading it, this is not good,” she said. “She’s a better writer than that.”
Larios Dominguez resubmitted her own essay, which she also finished before the assigned deadline, after The Hook followed up with Prater.
“I kind of knew from the beginning that [Prater] was gonna know it was me,” said Larios Dominguez. “It gave a more generalized essay instead of answering the question.”
Prater agreed that the AI-written essay was very vague and nonspecific compared to Larios Dominguez’s essay.
“It didn’t answer the prompts.” said Prater. “It was a summary and not an analysis.”
She advised English students to refrain from using ChatGPT for assignments.
“English teachers are looking for specific things in your writing, and if you don’t have all those things, you won’t get the points.”
Graphic by Dunya Omar
