From Wadsworth to Coventry to Green, school districts ask: Is ChatGPT cheating or a tool?

Local school districts are struggling to devise policies regarding the student use of OpenAI's app ChatGPT and other chatbots.
Local school districts are struggling to devise policies regarding the student use of OpenAI's app ChatGPT and other chatbots.

It wasn't long after the release of ChatGPT on Nov. 30 that reports of students using the program to cheat hit the news, and they continue.

On March 20, for instance, a New York Post article detailed a ChatGPT cheating scandal at an elite Florida school. Several students, the paper reported, admitted to submitting ChatGPT and other artificial intelligence programs' answers as their own.

At least one survey of students has shown that most believe using a chatbot response as one's own is cheating. But the same survey shows students will often do it anyway.

ChatGPT uses AI to generate everything from essays to term papers.

To respond to concerns, Open.ai, the creator of the program, released a tool to detect if a text, or part of a text, was written by ChatGPT.

The program, however, is not infallible.

University of Akron history professor Michael F. Graham said he has been able to detect at least one probable misuse of ChatGPT by students simply by asking the program if it wrote a selected block of text.

Graham said when he became aware of ChatGPT, he investigated its capabilities.

"I thought, 'This is probably something I need to make myself aware of,' " he said.

He found the program is capable of writing "a boring B paper," will only occasionally cite its sources and can make factual errors.

"ChatGPT also plagiarizes," he said. "(It) doesn't use quotes or citations."

Students  at Stonewall Elementary in Lexington, Ky., work on an assignment to figure out which writing sample was created by the artificial intelligence tool ChatGPT, which can generate everything from essays and haikus to term papers in a matter of seconds.
Students at Stonewall Elementary in Lexington, Ky., work on an assignment to figure out which writing sample was created by the artificial intelligence tool ChatGPT, which can generate everything from essays and haikus to term papers in a matter of seconds.

Still, Graham said the technology will have a dramatic effect on education, comparing it, in some respects, to the introduction of the printing press and its effect on the societal landscape.

He said students in one of his courses had discovered productive uses for the chatbot that don't spark ethical concerns. For instance, ChatGPT is capable of translating texts written in a difficult Scottish dialect.

"They found you can take a block of text, place it into ChatGPT, and translate (it) to modern English," he said.

Graham said he's aware the technology can be abused, but said at its current stage of development, instructors at the college and high school level have tools to detect cheating.

Graham said knowing a student's writing style is a first line of defense.

"As I get to know students, I get to know their writing voices as well," he said.

A ChatGPT-produced response would vary from the style of previous submissions and be detectable by teachers familiar with students, Graham and Barberton City Schools Superintendent Jeffrey Ramnytz said in separate interviews.

"We don’t have a policy just yet," Ramnytz said. "We will eventually sit down and figure out something if we need to."

The history professor said he suspects high school teachers will face the greatest challenges with increased chatbot usage.

"At the high school level this is really going to put a lot of pressure … on teachers," he said.

Don't ask ChatGPT to write an accurate biography

Kent State University associate professor Lucas Engelhardt, who teaches economics at the university's Stark campus, recently explored the limitations of ChatGPT and posted about the experience in a public Facebook post.Engelhardt asked the chatbot about himself and received a slew of inaccurate information in the response.

The chatbot was wrong about where he had earned his doctorate in economics, his specialization, where his work has been published and the awards he'd received.

ChatGPT did get some things right, Engelhardt said.

"(Long) story short, ChatGPT has confused me with someone much more impressive than I am …," Engelhardt wrote.

How teachers feel about ChatGPT in the classroom

A survey of 200 K-12 teachers by study.com two months into the ChatGPT era found that 80% had become aware of the technology and 26% had caught a student cheating with it.

Almost half — 43% — of the teachers surveyed thought it would make their job more difficult, but two-thirds did not think the program should be banned.

The survey also suggested the newness of the program and students' rapid adoption of it had caught educators by surprise, with 72% of teachers saying they'd received no guidance from school administrators on the issue.

The surveyed teachers did see potential for productive ChatGPT usage, including 39% who think it could help develop critical-thinking skills. About one-third of the teachers responded that it could help with problem-solving, data analysis and studying.

In a Dec. 9 essay in The Atlantic online, California high school teacher Daniel Herman questions ChatGPT's possible effect on the learning process and learning itself."Teenagers have always found ways around doing the hard work of actual learning," Herman wrote.

Many, if not most, students are not interested in developing even the basics of clear and coherent writing, despite its perceived benefits, Herman argued. He questioned whether, in the context of ChatGPT and students' willingness to use it to avoid the steps necessary to hone writing skills, the process of teaching and learning to write will remain relevant.

"(If) most contemporary writing pedagogy is necessarily focused on helping students master the basics, what happens when a computer can do it for us?" Herman asked.

Some students, too, see the potential threat of chatbot technology.

In an online article in the Feb. 3 New York Times, students discuss the pros and cons of ChatGPT, offering their thoughts on its implications.

"One of my biggest worries is that I would rely too much on these tools and lose the capacity for critical and creative thought," a student from Illinois wrote.

How some Northeast Ohio school districts are responding to ChatGPT

In the short term, though, educators are struggling to devise strategies to limit cheating while incorporating ChatGPT into work assignments.

The key, local superintendents said, rests in how it's used.

"There are probably many ways that that type of system can be used in beneficial ways," said Andrew Hill, superintendent at Wadsworth City School District. "This is something that is going to be around our society. We need to teach how to use (it) responsibly."

Springfield Superintendent Shelley Monachino agreed, saying: "I believe they are not going away, so we are going to have to figure out on how to make (chatbots) a tool."

Some local districts, like Green Local School District, expect to implement guidance on ChatGPT use after a collaborative effort involving the district's leadership team and curriculum and technology personnel.

Julie McMahan, coordinator of communications, community outreach and student wellness at Green Local Schools, said in an email that the district is researching the technology with a policy in mind.

"New guidelines, policies, and staff training will be forthcoming to address AI …," she said.

Likewise, the Manchester school system doesn't have a formal policy but one is in the works, Superintendent Shaun Morgan.

Leaders of some local districts, including Coventry and Wadsworth, said they're waiting to get a better understanding about the technology before proceeding with any policies.

"Due to the novelty of this application, we are still discussing how to proceed from a policy standpoint," Coventry Superintendent George Fisk.

"We want to make time to fully understand what are dealing with," Wadsworth Superintendent Andy Hill agreed.

'Educators will innovate ways to use this new technology'

What we're dealing with, according to eccentric billionaire Elon Musk, is a long-term threat to the future of civilization. Musk, who owns Tesla and SpaceX, was involved with the early development of chatbot technology with OpenAI, but left the board in 2018, long before the release of ChatGPT.

At one point, Musk called the threat of artificial intelligence greater than that of nuclear weapons.

For now, however, ChatGPT is a bigger threat to creative writing and accurate essays.

Fisk, the Coventry superintendent, is optimistic about the potential and long-term effect of chatbot technology in schools.

"I do not feel these applications are a long-term threat to critical thinking or language skills," he said in an email. "Educators will innovate ways to use this new technology just as they have each time a new technology has emerged in the past."

Leave a message for Alan Ashworth at 330-996-3859 or email him at aashworth@gannett.com. Follow him on Twitter at @newsalanbeaconj.

This article originally appeared on Akron Beacon Journal: What does ChatGPT mean for education? Akron-area educators split