Cheating… An age old problem…
I’ve been teaching for 25 years now, and I’ve been catching students cheating for about 24 years.
When I was an undergrad, everyone knew who to go to get a copy of assignments from previous classes. In 2010 there was a story of 200 students who were caught/admitted to cheating at my old Alma mater by using a test bank they found online (source). (Want a good watch… watch the YouTube video of the professor chewing out his class.)
Professors have been lamenting students using AI for some time now. While it’s hard to count, some sources are saying 84% or more students are using generative AI to cheat… I mean help them… with assignments (source) over a given semester, and up to 54% use a generative AI tool for school work at least once a week. While TurnItIn is saying that about 10% of assignments are 20% AI generated or more. (source). (It is worth noting that AI detectors are notoriously bad at actually detecting AI generated content providing both false positives and false negatives on data submitted to them.)
Some even say it’s good that students are cheating using AI (source), while others blame it on professors not updating their practices and thus “encouraging” it since it’s so easy to get away with.
The first fall after ChatGPT was released to the public, I taught an Honors course at Tusculum University on Ethics and Technology. While I hadn’t originally planned on covering AI when I offered to teach the course in the early spring, that all changed over the summer.
During the semester, I had a student ask why professors were so upset with students using AI. So I asked the question in reverse. “Would you be OK if I used AI to teach you everything?” “Of course not,” she said. “I expect you to know what you’re talking about.” I stared at her for about two seconds while she realized what I was really asking. I wanted her to think, reason, and explain things on her own.
This wasn’t a “You won’t always have a calculator with you,” argument. (Who remembers their teachers telling them that?) This was a “I’m trying to teach you to think… to reason. If you off load that work to someone else, what are you? How can you truly define yourself?”
For a generation who is often called the anxious generation, often because they have a hard time defining themself, off loading the one thing that can truly define them is not the wisest decision in my mind. And to do so in an Ethics course, which requires one to be able to define not only what they believe, but why, it is even more important.
But… this doesn’t explain what it’s like for when the professor uses AI. They supposedly already no the source material, and know the value of the content.
So what happens when a professor is caught using generative AI?
This is an interesting question. One student at Northeastern University noticed that her professor was doing just what I talked about with my student. They were had copied prompted information directly from ChatGPT into their handouts and grading responses to the student(s). (source) Of course the professor didn’t allow for generative AI from his students.
This has led to at least one student demanding a refund of their course. They feel as if they didn’t pay to be taught by a computer. But what about the Bill Gates prediction that AI will replace teachers within 10 years (source)? Is this just a professor trying to stay relevant as much as possible? Is it due to lack of resources? Or did they just not want to do that part of the job, and so they off load it?
If you want me to be honest… I’m surprised it took this long for some professor to be caught doing this. We’ve already been catchings researchers, doctoral students, and professors using generative AI in published papers (source) with a researcher finding at least 60,000 articles to have used LLMs in their papers.
When is AI a tool? When is it cheating?
I think this is the real question.
Yes, AI can be a tool to help you get the job done much faster and in some case better. But at some point it quits being a tool, and is cheating. Whether it is cheating a student out of an education, or cheating a student by not having a professor teach them is irrelevant.
To demand that professors do all their work by hand is ridiculous. Hand writing notes, handouts, papers, without using a computer is foolish, and no one would reasonably suggest it. Likewise tools like spell and grammar check are invaluable (I’ve lost count of how many spelling mistakes have been caught so far in this article) but are versions of AI. They just are a full generative AI tool.
But what about someone who uses an editor? My wife proof-read most of my papers for my Master’s degree. Does that mean I didn’t do all of my own work, even if she only provided grammatical assistance?
But let’s go back to the professor who caught his students using a test bank. This means that the professor didn’t write his own questions. He also used a test bank, at least initially. Does that mean he cheated his students out of his skill? Or that he was unskilled in properly asking questions, so he had to “cheat” and use a tool provided to all professors who use that textbook?
How about a professor that uses a teaching assistant (TA) to help grade his assignments?
What is the professor’s job?
This becomes a secondary question. While we have a romanticized idea of professors knowing their students, speaking with them as they casually stroll around campus, and grading each paper carefully, this isn’t always the case. At large institutions you may simply have too many students to grade each one’s papers. Does using a TA count as cheating the students out of the full teaching experience? (Note: I worked as a TA who helped do grading while an undergrad… I never thought of it as cheating. Should I have?)
I have a friend who recently taught a intro to programming class at his university, similar to what I teach at mine. When I teach it, I typically have a class of 18 to 25 students at our small private college. He teaches at a large R1 (Research Level 1) state university (with approximately 36,000 students), and had a class of 350. I can get to know my students individually, he cannot. To his university, he is more valuable as a researcher where he has brought in millions of dollars in grants and been published in over 200 academic journals and spoken at conferences around the world.
I attended an R1 school as an undergrad. We boasted a great faculty, of which many students never saw as they were busy doing research. Several were up for tenure while I was a student. There were rumors that the school didn’t even look at a professor’s teaching reviews – only how much research money they had brought in and how many papers they had published.
As the school gave little to no value to teaching, there was a misaligned value system applied to what the students expected vs what they received. What many thought should be the most important part of being a professor, was relegated to second, third, or even lower on the priority list.
I work at a teaching institution where we have smaller classes and not the expectation of research at the level that a research heavy institute like many state colleges have. With 100 students or less per semester, I can get to know my students in ways that my peers are a large state school never could. This is part of what separates us from those schools, and it doesn’t limit the outcomes of our students. I have students who have gone on to get their masters, and one who is currently working on finishing his Ph.D., while others have found careers in industry.
Is AI just a tool when your job is on the line?
When your job is on the line, and the schools do not value a professor’s interaction with (undergraduate) students, I can understand why some professors are turning to AI, because in reality its not what their job is. AI becomes a tool of convenience.
This however doesn’t not making it right, or not cheating the student, at least in my book. The student’s expectations is that they are there to learn from the professor. I’m not sure exactly where to draw the definitive line as to where AI quits being a tool and starts to be cheating.
A student might see the summation of a long work as a tool. I see it as cheating as there is an important skill in learning to summarize. It helps you learn what is important and what you need to know. A professor might see leaving feedback on a students paper as proper use of AI, where I see it as cheating as it doesn’t allow them to see issues in the class that need to be recovered either with an individual or as a group.
How I use AI in and out of the Classroom
As a professor I openly admit to using AI to help do my job. In fact, I am open with using AI not only with my fellow professors, but also my students. I have worked with students on two different projects to help them learn about AI, what is can and cannot do from a software engineering perspective. But I’ve also used in in ways to help me do my job:
- writing emails
- rewriting emails I’ve written
- coming up with test questions (this is a skill I’m not good at)
- generating images to embed in my notes (like the one above)
- generating podcasts of my notes for students who want to listen instead of read
None of these effect my core responsibilities as a professor. Those core responsibilities include:
- Determining what topics to cover in a class
- Presenting information including creating class notes, handouts, etc
- Coming up with projects and assignments
- Grading assignments and leaving feedback
- Mentoring students
- and more…
While some might argue any use of AI is too much (remember spell check is a simplified version of AI – so be careful about your absolutes), I think it needs to be used as a tool. Maybe this means schools defining what professor’s can and cannot use AI for, or having to disclose whenever it is used.
It would be foolish to think we could answer the question in a simple blog post like this. But it is something we will need to do, and the sooner, the better.
For the record…
I didn’t use AI to write this… but I used a lot of spell check… and there is that one image… but you wouldn’t want me to paint that any more than you’d want to read this without me using spellcheck. Trust me.
If a Student uses AI is it Cheating? What happens when a Student Catches a Professor Using AI? was originally found on Access 2 Learn
