CHLOE KOWALYK
Managing Editor
Students today feel like they are living in a dystopia — a world in which artificial intelligence (AI) can write essays, explain the material and complete assignments.
Perhaps the most prominent form of AI hitting campuses today is ChatGPT: an artificial intelligence model that creates texts based on prompts. Typically, the AI’s creations appear as though they were written by humans.
ChatGPT was created in November 2022 by OpenAI. Since its conception, the program has continuously gained accuracy and knowledge in developing texts. Students have begun to use the software to write their papers and answer questions on assignments. But is AI going too far? And where does the line fall between plagiarism and fair use of resources?
These are just a few questions on the minds of those who’ve come into contact with ChatGPT.
Graphic by Ryan Ludu
Dr. Michel Kouadio is the chief information officer at SUNY Fredonia. Kouadio explained that ChatGPT is the “first major breakthrough of its kind in the consumer space of internet search,” as it is almost able to think and imagine new answers to things that may not be obvious or predetermined.
Kouadio said that computer systems are based solely on binary choices, which include yes or no responses. These outcomes tend to be predictable. ChatGPT’s ability to discover new answers and make decisions are what makes it unique.
“With ChatGPT, you will have a more conversational interaction with the system, as if you were interacting with a knowledgeable advisor,” Kouadio said. “One of the best ways to understand this is to just go online and simply type a simple keyword or request such as ‘SUNY Fredonia’ in both Google and Bing Chat (ChatGPT). The different results will be obvious.”
Many feel that it can be tricky to determine whether or not something was created by AI.
Kouadio said it is typically easier to find artificial language and inconsistencies in large bodies of text, but shorter messages might be a little harder to find discrepancies in.
“Using our human faculties/senses alone, I believe we may soon reach a limit of our ability to know whether something was created by AI, especially advanced AI,” he said.
ChatGPT’s ability to make decisions and create artificially intelligent text is the reason many students are turning to the program to help them write essays and complete assignments.
Dr. Jonathan Chausovsky is an associate professor in the department of Politics and International Affairs. His views are his own, and are not reflective of the views of SUNY Fredonia or his department.
Chausovsky said that the issue of people using AI software to complete assignments isn’t a new issue, and noted that ChatGPT only makes it easier.
Kouadio felt the same way, explaining that we have seen similar rises in the use of technology over time.
“Whenever there is a new technology, we always wonder about the implications, and sometimes, we simply exaggerate the consequences. As an example, when online education started, some people thought that would dilute and make college education meaningless,” Kouadio said. “Instead of trying to combat the negative effects of ChatGPT, I’d rather suggest we study it, understand it, and see how to leverage it or similar tools in our education.”
Students have not just been using their peers, but also paying online writers a fee and using the internet to create essays and complete assignments for them.
Chausovsky said that with how easy it is for students to use ChatGPT, “the temptation is there.”
Chausovsky isn’t sure if he’s come into contact with any assignments written by ChatGPT in his classes. However, he has reduced the number of papers and is instead having his students take exams. While he said doesn’t prefer this, it helps to mitigate the chance ChatGPT is used.
Dr. Michael Sheehan, an assistant professor in English, also isn’t sure if he has seen ChatGPT in the classroom.
However, he does see the use of ChatGPT for school work as way to “rethink what the typical academic essay is for.”
For instance, Sheehan said that essays are typically used to view how students think and form arguments. “I think many of us are already looking for that in the writing we assign: don’t list facts or outline claims; show me how you think, show me the particular ways in which your mind navigates the world,” he said.
Chausovsky said, “I’m not convinced that the actual product of these things isn’t necessarily very good.” He explained that with ChatGPT, arguments don’t develop throughout the piece, and the piece is written, but “doesn’t actually say anything.” AI-created pieces typically lack a logical progression.
Kouadio said that new tools are being developed to help detect if something was created by AI. One such program is GLTR (http://gltr.io/).
Kouadio also provided some advice for students navigating a post-ChatGPT college experience.
“I’d suggest you strive to learn ChatGPT by running different queries,” he said. “Educate yourself online about how other students are using it. How far you should go will depend on your own ethics, at least for now.”
Chausovsky said that by using ChatGPT, students will not be developing the skills they need to think critically and form arguments. Doing work independently will allow students to fine-tune their skills.
Kouadio has a few thoughts on how ChatGPT will develop in the future.
Much like nuclear technology and computing/information technology, Kouadio said, “There may be misuse, as was and is the case with many other technologies,” he said. “I am personally very optimistic as an individual. So I like to think that most people understand their responsibilities in many things.”
Chausovsky compared the rise of AI to the introduction of social media. While social media seemed to be a positive change, it ended up having destructive consequences.
Sheehan said, “Although we are not wrong to be worried, we can also choose whether we want AI to do these tasks for us. Maybe there is a way to adapt, to work with what these models can do. But maybe the fact that we could unleash mass misinformation should give us pause.”
“Of course, [ChatGPT] may cause disruptions in many areas: jobs, education, economy, security/defense, etc,” Kouadio said. “Still, when all is said and done, most likely, we as a society will figure out how best to use AI, and that will be used for the greater good.”
Note: The headline of this article was written by ChatGPT. Could you tell?