Two things are infinite: the universe and human stupidity; and I'm not sure about the universe.
Albert Einstein
Trill News
Arts Culture STEM Competition Friday 27th December 2024 Industry Opinion Local Nations

Professor's Misuse of AI Software Causes Diploma Delay for Texas A&M University-Commerce Graduates

2023

A group of seniors at Texas A&M University–Commerce, who had already celebrated their graduation, found themselves facing a temporary delay in receiving their diplomas. The cause? A professor's misguided attempt to detect cheating by using AI software to assess their final assignments. However, the professor's methodology was flawed, and the incident has shed light on the need for educators to have a better understanding of the technology they employ.

According to the partner of one of the affected students, identified as DearKick on Reddit, Dr. Jared Mumm, an agricultural instructor and campus rodeo teacher, sent an email to the students explaining that they would be receiving an "X" grade in the course because he had used an AI chatbot called "ChatGTP" to determine if they had used the software to write their papers. To everyone's surprise, the bot claimed to have authored every single assignment.

There was just one problem – ChatGPT, the actual name of the OpenAI chatbot, does not function as a tool for detecting AI-generated content. It is not designed to identify material produced by itself or any other AI. In fact, the bot is known to sometimes produce misleading information, and it can even claim to have written passages from famous novels like Crime and Punishment. Educators have a range of effective AI and plagiarism detection tools at their disposal, such as Winston AI and Content at Scale, but ChatGPT is not one of them. Even OpenAI's own tool for determining if a text was written by a bot has been deemed "not very accurate" by a digital marketing agency.

Despite the students' attempts to prove their innocence, including providing timestamps on their Google Documents as evidence that they had not used ChatGPT, Mumm initially dismissed their claims. Frustrated, the students reached out to the dean and president of the university to seek assistance. Some even met with administrators in person to plead their case. It is alleged that Mumm handed out failing grades to multiple classes using the same flawed method, rather than questioning the validity of his approach.

In an ironic twist, a Reddit user named Delicious_Village112 conducted an experiment using ChatGPT. They submitted a section of Mumm's own doctoral dissertation on pig farming to the bot and asked if it could have generated the paragraph. Surprisingly, the chatbot responded affirmatively, stating that it could have produced the text given the right prompt. When asked if it had written Mumm's email to the students, the bot responded positively as well, but clarified that it would not have a record of such an email if someone had used its abilities to help draft it.

DearKick expressed hope that their fiancée's meeting with the university's Dean of Agricultural Science would resolve the issue. They speculated that Mumm may have had little familiarity with chatbots before attempting to use one in this manner. However, the situation became more complicated when two students admitted to using ChatGPT earlier in the semester, potentially complicating matters for those who had not.

In response to the incident, Texas A&M University-Commerce released a statement indicating that they were investigating the matter and developing policies related to the use of AI in the classroom. The university denied that any student had failed the class or been prevented from graduating as a result of this incident. They clarified that some students had received a temporary grade of "X," indicating "incomplete," to allow time for further examination of whether AI was used in their assignments.

The university acknowledged the need to adopt AI detection tools and other resources to manage the intersection of AI technology and higher education, recognizing that the use of AI in coursework presents a rapidly evolving challenge for all learning institutions.

This incident serves as a reminder that while educators may have valid concerns about students using AI for assistance in completing assignments, they also need a solid understanding of the technology to effectively address and prevent cheating. Ideally, this situation will serve as a learning opportunity for everyone involved.

Privacy Policy Contact Us