Texas colleges forced to embrace ChatGPT while addressing plagiarism concerns
The University of Texas, University of Houston and Austin Community College are finding ways to integrate the app into coursework instead of banning it.
Some Texas universities and colleges are adjusting coursework and policies to embrace the growing popularity of ChatGPT amid plagiarism concerns.
This new wave of advanced artificial intelligence (AI) can be academia's best friend or worst enemy. There are a lot of options, but ChatGPT is now the fastest growing app of all time and it's easy to use.
Ask it almost anything and you will get what seems like a well-thought-out answer. The chatbot can have a conversation, create art or even write an essay.
ChatGPT became available to the public in November 2022, but it's already taking the world by storm. No matter how old you are, the intelligence of cutting-edge technology is shocking.
"I'm 40," Chase Thayer said. "It is this most amazing tool."
"I think it's, like, crazy," University of Texas sophomore Jordan Tunnell said.
What universities are doing
The new Generative Pre-trained Transformer, or GPT, has already made its way into the academic system.
One student said he sometimes uses it for his math homework. Even while on the UT campus getting video to tell this story, we found a club using ChatGPT for coding.
"This gives you a nice first draft of some fresh idea, like ideas to use in your code," senior Tushar Kohli said.
The innovative technology has colleges and universities reevaluating lesson plans. UT Vice Provost and psychology professor Art Markman said he is excited about what he calls a tool. But he made it clear to his students that using it for unauthorized coursework is cheating after the bot did pretty well on an exam.
"I gave it questions I might ask on an undergraduate exam in cognitive science, and you know what? It did OK," Markman said.
Markman said UT does not plan on changing honor code policies but instead will give professors the option to ban ChatGPT or utilize it in coursework.
"It's already been improper to have somebody else write your essay," Markman said. "This somebody else is just an AI system, and so it still violates all of the policies that we have on the books."
University of Houston math professor and Vice Provost Dr. Jeff Morgan said the university doesn't plan on changing its policies either.
"There was a time when faculty members didn't want their students going online to get information, right? That's all past," Morgan said.
Austin Community College, on the other hand, is looking into updating its academic integrity policy.
"There will have to be at least some updates to [our] academic integrity policy, some greater definitions, perhaps helping students understand what plagiarism is, what collusion is, how to cite sources appropriately," ACC Vice Chancellor Gaye Lynn Scott said.
Using ChatGPT in coursework
All three higher education system spokespersons said one thing is for sure: they cannot outright ban the use of ChatGPT and other AI like it.
Morgan said the chatbot is brilliant, but it doesn't know everything.
"It took my midterm and my sophomore-level linear algebra class, and it got about a 50%," Morgan said.
Morgan is using that flaw to incorporate ChatGPT into his linear algebra coursework.
"I'm going to ask the questions and get answers from ChatGPT," Morgan said. "Instead of asking students to solve these problems, I'm going to ask them to analyze what ChatGPT has given as answers and tell me whether those answers are correct or not and why."
Scott said some professors at ACC said one way to address plagiarism is to have students write in the classroom more. But there are also ways to use the app to improve writing.
"You could bring an AI-generated essay into the room and ask students to make it better, to critique it, to find its strengths and weaknesses," Scott said.
When signing up, Open AI, which invented the bot, warns you it occasionally generates incorrect, misleading or biased content.
"ChatGPT doesn't have that reflective capacity, so it doesn't actually recognize that it said one thing and then, a minute later, it wrote something that was completely the opposite of that," Markman said.
Markman said the app learning from a large data set and incorporates feedback from users over time. So its creation of text, images and videos will only improve – but Markman said he isn't too focused on students plagiarizing using the chatbot.
"My feeling is that at the point where you're getting into an arms race with students about, you know, cheating and catching people cheating, you've kind of missed the point," Markman said. "That's why I really do come back to this idea of community. This is what this class is about. This is why it's important that you do the work."
The future of generative AI
Experts said ChatGPT and other technology like it will be as revolutionary as the calculator or spell check, if not more. To keep up, Scott said ACC is rethinking the certifications it offers.
"If it can do some of this basic work, we might need to not offer a certificate in what, press release writing or something, right?" Scott said. "The kinds of jobs we thought we were training students for three or four years ago don't exist. So I think it has more to do with how we pivot as an institution to make sure that we're helping students learn for the jobs that are going to be out there in the year or two or three down the road."
The three education systems are testing tools that could possibly detect if a student is using AI, but the technology is not 100% accurate yet. They've also created committees to address advanced AI or are planning training conferences to find ways to make sure students learn while using this fast-growing tool.
Right now, AI is unregulated. The Biden Administration is asking the public for input on policies that would hold artificial intelligence systems accountable.
Markman said it won't be easy regulating AI.
"I think they haven't completely figured out what the big-use case is yet," Markman said. "So unlike, for example, self-driving cars, where it's pretty clear what the domain is, I think with some of these large language models, it's a little bit less clear. And so until we understand what the market is, it's a little bit hard to regulate them."
With only a few months since its creation, one UT student said, "I think it already has changed the world."