AI in Education: will chatbots help or hurt teachers?

Maddie Genoways, Staff Writer

“Over the years, artificial intelligence has made huge advancements, from simple rule-based systems to complex learning models capable of performing tasks previously thought to be only within the realm of human capabilities.” (ChatGPT). 

At first glance, that sentence looks pretty unremarkable. Most people would scan right past without realizing it wasn’t written by a human. With modern advancements of AI, it’s becoming difficult to distinguish between the man-made and the artificially-generated. 

One of the more recent advancements in modern AI is ChatGPT, a chatbot developed by tech research company OpenAI released to the public in November of 2022. Like most chatbots, ChatGPT’s main function is to answer questions and deliver answers in a conversational manner.  

According to OpenAI, what makes this program different is the expansive and adaptive database it pulls information from, allowing it to form more coherent and authoritative replies. In its current state, ChatGPT can execute commands ranging from defining terms and generating project ideas to writing a full academic paper and even denying the possibility of robot world domination. Big Tech companies like Microsoft and Google are keeping a close eye on the development of OpenAI. 

OpenAI claims ChatGPT was created as a more conversational form of chatbot capable of more developed interactions than its predecessors.  

“The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests,” OpenAI said on the software’s opening menu.  

Of course, ChatGPT isn’t perfect. Tim Garvin, professor and chair of the Computer Technology and Information Systems Department (CTIS), warns not to put too much faith in the program while it’s in its earliest stages.   

“ChatGPT is still a work in progress, it’s very likely to make mistakes,” Garvin said. “The AI works a lot like the human mind: it takes in feedback and uses it to improve its own database. Given the same length and opportunity as a human, the AI can expand its knowledge and gain a better understanding of what users want to hear.” 

At its heart, ChatGPT is still a computer with a limited data bank to pull from. While it can formulate its own answers, it needs human input to get it started.  

One of the main concerns with chatbot programs like ChatGPT is that they pull from databases of preexisting information written by humans that is not credited in the end product. Because of this, there’s a grey area surrounding the use of chatbots to generate original or academic text.  

While ChatGPT in its current state cannot produce wholly original content, several reports from the Harvard Business review found that it could still prove to be a very useful tool, especially in a field entirely centered around learning and questioning. The reports claim that in universities where AI is used to answer student’s frequently asked questions gives teachers more time to cover essential topics in class.  

Staffordshire University in the UK and Georgia Tech in the U.S. have rolled out chatbots that offer 24/7 answers to students’ most frequently asked questions,” Harvard Business Review said. “These tests have confirmed that many repetitive tasks and routines could benefit from the assistance of AI-enabled systems.” For these universities, the use of AI in the classroom was a useful tool for students to personalize their classroom experience and allowed professors to focus on more hands-on learning strategies.  

According to another study at the University of Queensland published in Times Higher Education, ChatGPT has the capability to condense the repetitive, daily tasks of education into just a few clicks. In the study, the program created feasible lesson plans, produced accurate summaries of academic texts, and provided example sources for research. 

“Certainly, there are areas where AI can be used as a supplement in the classroom,” Garvin said. “AI is already built into so many facets of our lives that it becomes almost transparent to us.” Indeed, AI programs are already being used in schools around the world to help automate systems that cut down on teacher and administrator workloads. Schedule planning, course catalogs, and even WSC’s own website guide,“Willy,” are left in the hands of computers after initial human input.  

While AI use in school structure is fairly common, ChatGPT is one of the first really powerful AI programs easily accessible to any student with an internet connection and an email address. This has some educators worried about potential student misuse of the technology to get out of writing assignments themselves.  

For one professor, the program hailed as a revolutionary tool has turned her classroom into a waking nightmare.  

Donna Smith, an English professor at a southern university, has had several experiences with students using ChatGPT and similar chatbots to forge writing assignments. For context, Smith’s name has been changed, as she is still under the employ of this university – and her story could potentially reveal the identities of her students.  

“My students were handing in a several-week-long assignment over footnotes when I noticed a few of the papers had some really weird recurring problems,” Smith said. “I was concerned that these students weren’t understanding the lesson, so I asked them to stay after class for a discussion. When that didn’t get us anywhere, I ended up looking through their papers over the weekend; something else was off.” 

After some advice from a friend, Smith began to believe that AI was used to write the faulty papers, and ran them through GPTZero, an AI detector that identifies certain styles of prose commonly used by chatbots. “According to the detector, all but one of their entries was AI-generated,” Smith said. “The AI had mixed up information and miscited sources almost constantly. Some information from a 1912 newspaper was cited as coming from a 1940s journal, stuff that just didn’t make sense when you read it all together.” 

While Smith had hoped to resolve the issue with the students quickly and move on with her class, the university had other plans.  

“When I brought the issue up with university officials, they were very concerned, not with the students, but with the possibility that teachers might be making false accusations,” Smith said. “I had to spend a lot of time outside of work trying to build my case, the whole thing was a massive stress and a drag on my ability to teach. I was shocked that the university was more concerned with my accusations than anything else.” 

When presented with Smith’s evidence, all three students confessed to using AI to write their papers, and were given punishments by the university. Because of this experience, Smith has come to view chatbots as nothing but a problem, saying “I wish it had never been invented, it has done nothing but make my life difficult.”  

While Smith acknowledges the potential applications AI like ChatGPT may have in longform, multi-semester courses, she maintains that they will not have a place in her classes.  

“I’m teaching single-semester basic writing and college speaking. My students have to try and stumble in order to make progress. They need to show me all the warts and scars in order for them to be made clean,” Smith said. “When you’re trying to teach young people to write, this is nothing but a huge frustration. I wish my students understood that this is a serious issue of plagiarism.” 

It’s still too early in the development and widespread use of the AI to determine whether it’s a friend or foe to the school system, but there are certainly plenty of ways for educators to make use of the potential new tool.  

“Since ChatGPT functions like an automatic FAQ, it can be used to quickly answer student questions and automate processes not key to learning,” Garvin said.  

On the other hand, there are just as many ways for students to abuse AI and make educators’ lives hell. Smith is not alone in her experience, but with accusations of intentional plagiarism and reputations on the line, it can be difficult to have an open, honest discussion about long-term student use of AI. 

While the AI is powerful, it is still a developing technology that makes just as many mistakes as it does insightful answers, especially when it comes to the creative or philosophical.  

“Don’t ascribe too much capability to the AI until you’ve tried it,” Garvin warns. “It’s easy to watch a five minute demo and think you’ve found the all powerful AI that will control the world. It should not be a replacement for human creativity, but as a tool that we can use as an extension of our own abilities.”  

On the whole, it seems the most common consensus about AI in education is that if it is to be used, it should be a tool secondary to human research. ChatGPT is here, and it’s only going to become more and more prevalent among students as the software improves. Every day, the world of AI is developing and growing, and it’s up to educators to decide how they will adapt to the coming changes.