This article previously appeared in the October 2024 edition of The Commuter.
If higher education were to dress up for Halloween, it would probably be as ChatGPT. The threat of artificial intelligence has rapidly become the boogeyman of countless industries, but perhaps none more than in education, where AI threatens not just job security but also academic integrity.
AI might scare faculty, but it’s a worry for students as well. The 2024 Digital Education Council Global AI Student Survey reported that over half of students polled believed an overuse of AI devalues education, although 86% admitted to “regularly” using the technology.
ChatGPT and other artificial intelligence tools have become prevalent to the point where they can’t be ignored any longer. Following inservice days ahead of the 2023-24 school year, Vice President for Academic and Student Affairs Ann Buchele launched an AI task force at Oregon’s Linn-Benton Community College. It first met in March of 2024, co-chaired by manager for Student Conduct and Retention Jill Childress and counselor and advisor faculty Sonya James.
The task force is set to run until spring of 2026, with goals to create school policy on AI, offer training for faculty, create a library of resources, and “launch a community of practice for faculty to be working and supporting each other through what AI use looks like from a pedagogical and curriculum perspective,” said Childress.
Childress presented an update on the state of the AI task force during inservice ahead of the 2024-25 school year.
The 15-person team has yet to make significant updates to the school’s reaction to AI, but on a topic that complex, time is needed. “The technology is changing daily … we want to make sure our recommendations are not too rigid,” explained Childress.
A number of basic AI recommendations were given out during the inservice session. Instructors were urged to have AI policies in their class syllabi, with the task force offering guidance and examples to those who were unsure where to start. There were also warnings about the fallibility of AI-checking programs, which are seen as useful tools but not recommended to solely determine whether a student’s work was AI-generated.
But the AI problem involves far more than just academic integrity. Childress highlighted equity concerns, ethical and security implications, environmental impact, and community and workforce applications as factors the task force has needed to consider.
For instance, some companies have begun to look for candidates with AI competency, something the school needs to consider when preparing students for the workforce. And because some programs such as ChatGPT learn from user input, instructors were warned of potential FERPA violations regarding entering personally identifiable information into AI tools. The task force said to use programs that cannot connect to the school’s database.
Although individual instructors are still open to make their own class policies, completely banning AI was not one of the options for LB’s task force.
“Using AI as a tool is the direction we’re headed in,” said Childress at inservice, noting how the task force itself was using the technology to organize its goals and synthesize data. With AI now readily accessible and built into many web-based educational tools, it’s getting harder for faculty to treat their coursework like AI never existed in the first place.
“I’ve been really surprised at just how useful it is as a tool,” said Childress. “I think just the iteration of it that we’re seeing now – there’s a lot that it’s capable of doing in ways I think can really save us a lot of time and then free us up for maybe more complex work or more sophisticated thinking.”
Childress and the rest of LB’s task force recognize the many pitfalls of AI, but the end goal remains to implement it into the school’s structure before higher education falls too far behind in the use of the rapidly developing technology. While still in its infancy, the task force is the first step towards LB entering the future of academics, the workforce, and beyond.
However, most Linn-Benton instructors have been dealing with AI long before the task force was formed. The consensus seemed to be that 2022 – when ChatGPT updated from version 3 to 3.5 – is when the first wave of generative technology hit the classroom.
“A couple years ago is when I first started seeing it really hit big,” said Joseph Jess, a former computer science instructor at LB now at Oregon State. “And of course, it’s all over the place now. It’s one of the fastest-growing things that we’ve seen, definitely one of the fastest-growing things I’ve seen in influencing student learning processes.”
Most faculty have had to adapt to the new technology, although AI stances vary with each class and instructor.
I want students to use AI as a tool and not as a crutch.
Mark Urista, a communications instructor, flat-out encourages students to experiment with AI as a tool for brainstorming and debate practice in his argument and critical discourse class. However, he recognizes that a course where most assignments are delivered orally has more built-in protection from AI than an online writing class.
“I encourage students to use AI to develop arguments for and against the proposition that they have been assigned, and then they can begin to analyze how strong those arguments are,” said Urista, adding, “You have to consider the class and the context in which AI would be used. I really feel for faculty that teach writing classes or classes where calculations can be done very quickly with AI.”
“I want to teach students how to use AI well, but that’s something I’m still trying to figure out,” he said. “Bottom line is I want students to use AI as a tool and not as a crutch. Students are in college to learn and to develop the ability to think. I don’t want AI to do the thinking for them, but I recognize that AI can be a tool that helps them with learning and thinking.”
In computer science, Jess has a similar stance: “My policy is basically, if I ask for some requirement in an assignment, I want you to have written that. If it’s anything else, I don’t care where it came from, but I would like it to be cited.”
Jess went on to state how using external sources has never been uncommon in coding, although in the professional world, uncited code can be a headache. It’s made worse by the unclear origins of some AI-generated work.
“I want [students] to start to gain an intuition of an ethic or something like that. … If they become a software engineer, they’re going to have to state exactly where something came from,” Jess said.
Lewis Franklin, the program chair of LB’s Visual Communications program, is able to go without an AI policy because he works with students throughout their entire creative process. “If you’re very careful about it and you only use it as an aid instead of a solution, you’re using it the way it should be used,” he said.
However, licensing is an important factor to keep in mind. While Adobe and other visual software have programs that generate from only licensed materials, others do not. Said Franklin, “I try to train my students as best I can to be cognizant that they’re working for clients. Sometimes there’s millions of dollars at stake, and you can’t play this game of borrowing from non-reputable sources.”
Anthropology instructor Lauren Visconti devotes around two pages of her class syllabus to AI usage, including a bulleted list outlining when it is and isn’t permitted:
“For example, the following learning activities permit the use of generative AI tools (e.g. ChatGPT, Bing, Gamma, Claude, etc.)
- Brainstorming and refining your ideas
- Finding information on your topic
- Drafting an outline to organize your thoughts
- Checking grammar and style
The use of generative AI tools is not permitted in this course for the following activities:
- Writing a draft of a writing assignment
- Writing entire sentences, paragraphs, or papers to complete class assignments
- Generating Powerpoints, slideshows as assignments”
Visconti also uses AI as a tool in her classroom at times. At one point, she wrote “What does it mean to be a human?” into ChatGPT for the class to analyze. “I think that could be a positive in bringing critical thinking skills around technology,” she said.
While AI tools can be a powerful brainstorming tool, anthropology instructor Sarah Walker added that they are still far from perfect, often amplifying the shortcomings of internet search results.
“I’ve used it to show students hegemony in history teaching and the teaching of what indigenous people are and are doing,” she said. “Because if you go into ChatGPT and you ask questions, like, ‘Who are the Mayan people?’ I’ve had students look at it and then basically go through and say, ‘Oh, well, this isn’t true.’ And so it’s reflecting a lot of the biases that are already within the culture. It’s a computer [program] written by people, so it’s going to have all those problems.”
ChatGPT can also struggle to keep up with changes in culture and information.
“You have the data on Monday, and then something drastically changes because it’s humans and culture on Tuesday – ChatGPT is going to give you the same s— on Wednesday that it gave you Monday,” added Visconti. “It does not matter what happened on Tuesday. And that’s scary, because it can prohibit cultural change.”
Some bad uses of AI – I would say for the most part, it’s people who run out of time.
Journalism and writing instructor Suzi Steffen’s classes follow a similar policy to Visconti’s: AI is allowed for developing ideas, but student work must be completed by students. However, in Steffen’s case, the use of AI is reluctantly permitted.
“I prefer people not do it just because of the environmental impact. But I mean, for brainstorming, what are you going to do? I mean, people use it. That happens,” she said. “I have my own personal issues around it, because I feel for authors and artists, and I feel if we have great, amazing machine learning, why aren’t we using it for, like, farm work – I don’t know, cleaning toilets. Like, why are we using robotics and large data sets for really great things?”
“So I don’t outlaw the use of AI – just when people turn in papers, I want those to be written by them,” she said. “And if they’re stressed out, I want them to come to me and say, ‘Suzi, I can’t do it this week.’ Like, OK, let’s talk. That’s a level of trust you have to have in your professor. Me, I have to trust my students.”
“What I always say is, look, if it looks like AI wrote this – maybe you weren’t using ChatGPT, maybe you weren’t using OpenAI, I don’t know, but it looks like it was, so just revise it. And please don’t complain to me. Just go, ‘OK, I’m going to revise this so it sounds more like me now,’” she said.
“And sometimes that means that people end up with some kind of unpolished writing. You know, people are between their freshman and senior years of college – you’re not going to be good at everything, not going to be perfect. You just have to get better as you go.”
Steffen outlines a forgiving stance many instructors are taking with AI cheating incidents. With the stress of college and the increasing accessibility around AI, many students are being given second chances to submit their own, original work.
“Some bad uses of AI – I would say for the most part, it’s people who run out of time,” said Jess. “They just ask a question of ChatGPT or other AI, and it spits something out. It kind of sort of works. They’re hoping for partial credit, right? … And I’ve had more and more of those as the past couple years have gone on, but my policy has remained roughly the same, which is, I will have a conversation with you as the student before I report you, because it’s entirely possible that maybe you just ran out of time, and you’re still willing to put in the time to do it correctly.”
“And so there were several times, a couple times per term, while I was at LB where I would have that conversation with a student, and nine out of 10 of them would say, ‘Can you just give me the weekend to try and remake this?’” said Jess. “Show me what you got at the end of the weekend. If it looks like you’re legitimately trying, you’re good to go. And I know some people have really strict policies on that kind of thing, oppressive policies, I might say. It does not foster learning to just say, ‘Get out of my class.’”
Jess also highlighted ways where AI could be an improvement over the questionable online learning resources of the past, mentioning a newly-developed “rubber duck” program. Rubber duck debugging is a term for explaining code to someone or something — an inanimate object like a rubber duck – and discovering mistakes and solutions through the process of explaining every detail.
“They made this AI rubber duck that was trained to be sort of a Socratic method, asking questions to clarify student questions so they can then learn. And this thing can be available 24/7, right?” said Jess. “Unlike a TA or an undergraduate mentor or your instructors or professors, it’s available in the evenings. It’s available on the weekends, and it’s trained in a way that doesn’t just write code for people. So it ends up still trying to teach them all of the basics of this thing without just throwing an answer at them like Course Hero or other Chegg sites might do, where it’s like, ‘Oh, we’ve already got an answer for that question.’ OK, how much are you going to learn by copying and pasting it, though?”
“I love the fact that AI can now serve as a tutor for students, particularly disadvantaged students who may not be able to access a tutor on their own,” said Urista. “I also love the fact that AI can provide something like a second reader to students. I remember back when I was in college, whenever I wrote a draft of a paper, there were lots of typos, lots of grammatical errors, and someone at the writing center could help me identify what those were and improve the paper. Well, now we have Grammarly.”
Walker stated how she wishes students would use AI: “Be transparent. Don’t try to be sneaky about it. Tell your professor, like, ‘Hey, I’m using AI for this, and it really helped me.’ You know, ‘I used AI for this, and it was just terrible and I don’t know why everybody thinks it’s so great!’ Whatever – we’re learning, too.”
I’m starting to ask for handwritten things.
However, there’s also an emphasis not to make it too easy for students to turn to AI tools to complete coursework.
“I put the onus of responsibility on the instructors,” said Urista. “Instructors need to figure out how to essentially make their class AI cheat proof.”
“One thing I would recommend that all instructors do is use AI to complete some of their assignments,” he added. “That’s probably the best way for them to get insights on what adjustments need to be made. … Once ChatGPT came on the scene, I started to complete [speech analysis assignments] with ChatGPT, and I realized AI can complete a good portion of this assignment without the student putting any original, critical thought of their own into it.
“Now, a lot of my papers require students to analyze speeches and debates that happen in class, specifically because I know AI cannot make connections between a lot of the content that we’re teaching and specific examples from those speeches in class. Making those connections are critical.”
“I’m starting to ask for handwritten things,” added Steffen.
But students may not all be as likely to use AI as some think. “One thing I’m hearing from my colleagues about students is that they’re not necessarily as on-board with AI as we might think right out the gate,” said Childress. “I mean, I’m sure plenty are, but there are still some reservations there, too.”
“I don’t think a ton of students actually use it,” said Alex Loveday, who was the Student Leadership Council representative on the AI task force.
Still, Urista notes that the younger generation seems to be more comfortable with the technology: “Something I’ve accepted is that students are always going to be one to two steps ahead of me. I’ve actually learned more about AI from my students than I have from any of my faculty or administrators or professional development training sessions that I’ve gone to.”
“One thing I would encourage my colleagues and other folks in higher ed to do is bring students into these conversations and let them know. I want you to speak honestly and authentically. Tell me the good, the bad, and the ugly. We need that kind of honest, transparent communication to occur if we’re really going to figure out how to chart forward in a productive way.”
Childress added that students can reach out to her or Sonya James in the Advising Center if they want to share their feedback or experience with AI at LB.
“We’re delighted to meet and hear from them, and get their perspectives, and then if there’s a lot of energy there, we can figure out ways that we can get them plugged into the group,” she said.
We’re all going to be replaced by AI at some point or another.
That feedback and dialogue could prove important as LBCC and other colleges plow into the unknown future of AI.
“We’re all going to be replaced by AI at some point or another,” said Franklin. “As AI starts to make more and more web pages and starts to do a lot of the other graphic design work, of course it’s going to have a hugely negative impact on what we do.”
“But the AI can’t sit down with clients and discuss things and figure out the human pieces when you’re dealing with branding and you’re dealing with graphic design, because you’re dealing with people, not with computers at that point,” he added. “And so I don’t know – I can’t imagine AI being able to replace that real interaction. You can replace a doctor with an AI, but what about bedside manner? You know, that human component that we need and crave. So, it’s a double-edged sword.”
Steffen was less confident in an eventual AI takeover. “I read someone talking about how all of the people who were into crypto are now into generative AI. And if you asked, like, eight years ago, what’s going to happen next with crypto, they would have been like, ‘It’s the most amazing thing ever.’ Well, I think that these are folks who live on their wits. They want venture capital. They want to be able to say, ‘I’m pivoting to this next best thing.’ So I’m not sure that it’s going to hold up. But who knows?”
Still, she warned about the effects of students over-relying on AI to eke out a degree: “What did those students learn? Well, they learned that if they were good enough [with AI], or their teacher was tired enough, or whatever – I mean, that is the way life works. But when we need somebody doing research to not cut a corner, for example, healthcare research or nuclear physics research or something, I don’t want that to happen.”
“I don’t know, maybe ‘frightened’ is too strong of a word,” said Visconti. “But email was supposed to reduce our work and time [laughs]. … So I’m skeptical until we are able to put labor at the forefront of a new technology.”
Childress has mixed feelings as well. “I guess I believe in the resiliency of humans, and so I think it’s easy to be a little afraid of what could come, but I’m hopeful that when we have people who are smart and capable and creative, that we will figure out ways to help guard against some of the potential pitfalls of generative AI,” she said.
“So I think I’m overall optimistic, and I’m hopeful that it’s a technology that can really free us up to do really cool, interesting work. … I’m interested to see where it goes – I think I’m in camp optimism.”
