Artificial intelligence: The way students complete assignments could change with new software that can do their work for them | News | San Luis Obispo

“This is a giant tidal wave coming to academia.”

That’s how Paso Robles High School English teacher Aaron Cantrell describes ChatGPT—a new artificial intelligence software that’s been the talk of classrooms all across the country. The software has ignited conversations among teachers at Paso High since its launch on Nov. 30, 2022. Created by San Francisco based company OpenAI, ChatGPT has the ability to formulate human-like written responses to any question or request asked by the user.

click to enlarge

  • Cover Image From Adobe Stock
  • HELP-ME Artificial intelligence softwares like ChatGPT can write essays, answer questions, and more for students.

Once Cantrell started exploring ChatGPT, he said he realized it was “a force to be reckoned with.”

“I asked it to write all of the essays that I’ve assigned to my students this year,” he said. “A good chunk of them did phenomenally well in 15 seconds on each of them. I said, ‘Wow, we’re gonna have to have a conversation about this, because it’s a bit of a game changer.”

The way it works is simple—ask ChatGPT to write anything, like an essay, an article, or even a cover letter. Within 15 seconds, ChatGPT delivers, which makes it fairly easy for students to cheat. A survey conducted by reveals that 89 percent of survey respondents said they have used the platform to help with a homework assignment.

Paso High School senior Cosmo Toohey-Bergvall explained that ChatGPT is blocked on the school’s Wi-Fi as well as students’ Google accounts, making it impossible for students to use the software on school-assigned Chromebooks even while they’re away from school.

Toohey-Bergvall said that the software lacks a “personal touch” when it comes to essays but noted that it could be good for “busy work” assignments.

“[It’s] a very, very useful tool that can create a large amount of text, push people off in a certain direction, inspire them, or just generally serve as a basis for your writing, rather than it being a thing that can create whatever you need from scratch from a few words and a prompt,” Toohey-Bergvall said. “It needs to be taken care of and nursed in the right direction.”

Cantrell said there could be a way to use the AI ​​software that facilitates actual learning.

“What can we do to maybe use some functions of AI to do research?” Cantrell said. “I don’t think there’s anything necessarily virtuous about going through a card catalog, or search engine, and finding things and putting them in order. Maybe that part of it we can outsource to machines. But then the accumulation of all these ideas and the driving of some kind of argument can be retained by human beings.”

Ryan Jenkins, associate professor of philosophy at Cal Poly, said that any integration of chatbots into education should be handled carefully.

“I think that that has a pretty significant potential to erode some of the values ​​of going through a college class. For example, if you reach the point—so, far end of the spectrum—where an AI is writing your essays for you, it’s not obviously different to me than having one of your classmates hand you an essay,” Jenkins said. “That is to say, neither of those is really challenging the students to reflect on their own beliefs to work through a different difficult puzzle.”

click to enlarge

NEXT GEN New artificial intelligence software could change the future of education, sparking debate around academic dishonesty.  - FILE PHOTO BY JAYSON MELLOM

  • File Photo By Jayson Mellom
  • NEXT GEN New artificial intelligence software could change the future of education, sparking debate around academic dishonesty.

There’s not a lot of plagiarism software equipped to deal with AI-written essays. So far, the most popular detection software is called GPTZero, which was developed by 22-year-old Princeton student Edward Tian.

Since the technology is so new, teachers and administrators at Cal Poly are divided on how to tackle the use of ChatGPT.

“I think you see a range of all kinds of responses from this sort of Chicken Little ‘the sky is falling’ response to people who are openly embracing it,” Jenkins said. “In the middle, I think you have a lot of folks that are saying, ‘Look, we can’t fight against this. You know, we can’t prevent students. [from] remembering that this technology exists.”

While there has been no official directive from Cal Poly administrators regarding what to do with ChatGPT, Jenkins said that there’s been a lot of “hand-wringing” within departments.

However, Cal Poly computer science professor Franz Kurfess has begun using ChatGPT as a learning tool. In his Computer Support for Knowledge Management class, Kurfess encourages his students to compare their proposal with a version that ChatGPT generated.

“A few students already experimented with it, and the results were decidedly mixed. So some of them said they were actually impressed because the results that ChatGPT delivered were reasonable, not perfect, but then the students’ proposal probably also will not be perfect, Kurfess said. “Other students said it was practically unusable. And it’s too early to draw conclusions, but my suspicion is that the students who didn’t get good results had very technical topics.”

One student who experimented with ChatGPT was fourth-year Computer Engineering major Brett Gowling, who explored the chatbot’s capabilities and its shortcomings through a presentation he did for Kurfess’ class.

“I think the main problem that you have to avoid with students using this is the direct adoption of the AI ​​output as one’s own. And I think you can use, or you should be able to use, maybe an outline or structure that the bot. developed,” Gowling said. “But the text should be either modified significantly, to make it your own, or it should be direct quotes, if it’s going to remain unchanged, and you need to give credit then to the chatbot.”

In an effort to further facilitate conversations about the implications of ChatGPT in academia, Jenkins and his colleagues published a report on Jan. 30 about the norms of using and crediting AI for its contributions to scholarship.

“There have been several papers that have been co-authored by ChatGPT and other large language models. So we sort of drew up some principles about this and how to think about this from the perspective of a scholar,” Jenkins said. “I think next we’ll try to suggest some language for the administration to propagate down to students.”

While it might be too early to predict what the future of education might look like with the introduction of ChatGPT, Jenkins said that teaching methods are going to have to change.

“My worry is that this technology will sort of start at the boundaries and creep in to colonize or to displace more and more of the kinds of activities and assessments that we would have in a classroom that are supposed to really challenge people to think,” Jenkins said. Δ

Reach Staff Writer Shwetha Sundarrajan at [email protected]


Leave a Comment

Your email address will not be published. Required fields are marked *