I plugged the following prompt into ChatGPT: “write a four-sentence introduction to a blog post weighing the pros and cons of chat gpt and its implications for academia, including the impacts on academic integrity, how professors can detect plagiarism, and the potential for inaccurate results. ”
Here is ChatGPT’s result: “In the rapidly evolving landscape of academic discourse, the integration of ChatGPT into educational settings has sparked both excitement and concern. This blog post delves into the dichotomy of the pros and cons surrounding the use of ChatGPT in academia, scrutinizing its implications for academic integrity. As educators increasingly grapple with the challenge of detecting plagiarism in a world where advanced language models can generate content, we explore the tools and methods professors can employ to uphold academic honesty. Additionally, we scrutinize the potential for inaccuracies in results generated by ChatGPT, shedding light on the delicate balance between technological advancement and the reliability of academic inquiry.”
It doesn’t say exactly what I’d like it to, but I think that may be a good thing. We can see the almost confusing style of writing that ChatGPT uses with long, wordy sentences and 25 cent words where 5 cent words would be more accurate. We can also see how my voice does not come through in the writing, and we can see that ChatGPT does not churn out an entirely accurate interpretation of prompts. I plugged in “how professors can detect plagiarism,” which ChatGPT interpreted as “the tools and methods professors can employ to uphold academic honesty,” which is a similar concept but seems more scientific than I am in this blog post. Still, it’s a fairly decent piece of writing for a computer and a huge technological feat.
Certainly, as ChatGPT itself generated, there is “both excitement and concern” regarding AI, especially in regard to academia. In my experience as a student, I have used ChatGPT to summarize difficult articles to make sure I caught all the major points, only to find ChatGPT did not summarize the right article or just made something up entirely. I have also used ChatGPT to generate a list of relevant sources for me to look into, and then, after a very frustrating half hour searching for these sources, I realized ChatGPT literally generated them– They did not actually exist. However, ChatGPT has been exceptionally helpful in generating discussion questions for me to make sure I comprehended difficult articles, and it has also been helpful in my language learning endeavors as a kind of digital pen pal to write to in French. (I also think it generated a pretty good introduction paragraph for this blog post!)
In my experience as a professor, ChatGPT has been incredibly terrifying. As I’ve only been teaching for a few years, I don’t have the experience that other professors do to catch that kind of uncanny style of writing that ChatGPT uses, and there have been times when the phrasing in student essays were cyclical, repetitive, and strange, yet I cannot prove that ChatGPT (or another AI program) generated their essay. In “ChatGPT and the Rise of Generative AI: Threat to Academic Integrity,” Damian Okaibedi Eke notes that “for universities to preserve the current assessment methods based on written essays, there is a need to create a reliable tool that can detect AI generated texts. However, designing such a tool and incorporating it into effective or reliable assessment approaches will require a lot of funding and the support or buy-in of OpenAI or other creators of these language models. It may also take time to develop.” There are clearly these major restraints on us as educators and us as humans. Eke suggests rethinking essays as assessment anyway, but in freshmen writing courses, it seems difficult to do anything at all. Setting the learning agenda based on a student’s sample essays is absolutely useless if the essays are not a reflection of students’ abilities but of AI’s abilities.
Nina Begus works on developing technology and the humanities, and she said once (or probably many times, but I heard her say it once) that technology is developing rapidly whether the humanities are involved or not, so the humanities better get involved. That stuck with me. We can talk about how these forms of technology shouldn’t exist, but they do. So we had better come up with a way to face these technologies in a way that maintains integrity. I’m not sure how we can stress the importance of academic integrity to our students in a way that feels real when there is still no highly accurate way of detecting AI-written texts. Perhaps we could do what Val O’Bryan suggests: “instructors can have [students] use the tool at different stages of the writing process, stopping them along the way to discuss and think critically about the generated content and searching through databases to try to find credible sources that support the generated content.” This is an interesting idea, but I’m still not sure it would prevent undergraduate students with essays due in five courses from using an AI-generated essay when deadlines got tight. However, it seems to be a good idea to speak openly about AI in classrooms to help to explain the limitations to students.
Several tools have been created to try to catch AI-generated instances of plagiarism, so it is, hopefully, only a matter of time before these tools grow more and more accurate. In the meantime, it seems to me that the best we can do is aim for open communication about AI in our classrooms.
Hi Gina,
As a teacher, I understand your concern with students’ use of Chat GPT with bad intentions. Most schools and universities try to prepare qualified students and candidates for the world as they are done with college. The fact that some students choose to use Chat GPT for academic dishonesty is sad because we all know that they are not lying to us but they lie themselves mostly. A lot of instructors tend to turn on AI-detection software programs which help to find out if a student used Chat GPT for cheating purposes. However, as technology is advancing, I think AI tools will only get better each time and the fact that they are a threat to most of the norms of academic institutions is absolutely true. However, as a teacher, I would rather learn how to insert AI properly in my classroom instead of chasing after students and panicking every time I have to grade papers or assignments. I always tell my other co-workers to think about AI as a game changer and not a game ender because if we try hard not just independently but collectively as institutions, teachers can be trained on using AI in classrooms in beneficial ways. As long as teachers learn how to incorporate adequately AI in the classroom and students use it wisely I see more benefits of AI in the Classrooms.
Hi Gina!
First, I’d like to start off by saying I appreciate your perspective into the fears that educators have in experiencing ChatGPT. It is sometimes challenging to recognize the difficulty in how large developments like this technology cannot be simply implemented into curriculum, as they change entire learning experiences. Similar to the development of the calculator, curriculum had to be altered drastically to fit a new type of learning, The same can be said about ChatGPT, as now one must consider how this will affect assessments and essay writing. Have you experienced ChatGPT in your classroom or have students brought it up to you during class? I question how they feel about all of these changes, especially after reading how divided students are about its uses, as I assumed all students would be for this development.
Also, I love how you incorporated ChatGPT responses within your piece, I thought of doing it as well but could not pull it off as seamless as you did! I found it to benefit your argument as it shows the awkward language is easy to detect and separate from your writing style. As you said, we must be open to communication about AI in our classrooms, as these developments will continue.
That was a really smart post, and I loved how you began by putting one of our prompts into ChatGPT. What it spit back at you was really predictable, but not embarrassingly so–many, many students have been beginning their papers in such ways well before the chatbot boom. I felt for you when you said that as an inexperienced teacher you feel that you’re disadvantaged because you cannot detect the odd repetitive prose that seems to mark AI produced papers, but at the risk of scaring you more, I don’t think any of us are really very good at this. We can recognize certain red flags but I do not think they are necessarily a mark of AI writing–they tend to be marks of poor, unengaged or developed student writing. And some output of AI writing is like that. The students who will get caught are the ones who don’t take the time to work WITH the AI but have it work for them. Even those who do work with it may not understand how much they have to work beside it in order for it to be a useful tool rather than an actual problem for them academically and intellectually. As I’ve said in other responses, I think poor reading comprehension skills are going to be key in whether ChatGPT can be used as an effective learning tool or not in writing instruction. Your reading skills are what allowed you to see the flaws in otherwise good looking output. Not to mention your skill in double checking or triple checking the output if necessary. The superficial polish of ChatGPT is a shiny treat dangled in front of many people (including the scientists who used it to write an article for Nature–a lot of academics really do not like doing their own writing). I would tell students that it’s worse for them to let ChatGPT be the endgame (last stop) for anything regarding their own research than anything else–and that includes using it to write a paragraph or some sentences or a conclusion. It’s not so much that it’s trash as that it is not a one-stop shopping for anything intellectual.