AI Plagiarism Considerations Part 3: Having the AI Conversation
AI Plagiarism Considerations Part 3: Having the AI Conversation
We gotta talk about it...we just do...
Here’s the final part in the series as I promised. To recap, we first talked about AI plagiarism checkers then moved on to engaging with students you suspect of using AI. Today, we’re talking about “the talk.”
The talk is something you have early in the course. In truth, the talk can extend to academic dishonesty as a whole and a slew of other topics. It’s probably best to spend a good portion of the first class on a talk like this—establishing a dynamic of listening, learning, and build trust with the students. What follows are what I think are the key highlights or aspects of that talk. The order isn’t predefined and you can certainly interweave them as you see fit, but I do believe at least this much is needed.
Admit it exists and that’s ok
I remember early only some folks where were even afraid of mentioning it. There was a bit of the fear that they would induce the Streisand effect for their students and they would start using it (even more). But generative AI isn’t Voldemort and avoiding mentioning isn’t going to do much other than lead students to believe you’re not aware of it or that it’s fully allowed in the course because it doesn’t come up anywhere.
Open up the conversation about AI and be clear about some facts. Generative AI exists and there’s many things we don’t know about it as it changes and we come up with new and difference uses for it. It mimics human writing in many ways. It can do work that previously couldn’t be done by machines in such an automatic fashion. It can be incredibly useful in many ways. It can also be tempting to use it in lieu of work that is structured to hone your abilities—especially in courses that you may feel challenged by, inadequate to complete, or just have a disdain for.
That’s the truth and it’s what we need to talk with students about. And I know for many of us—this is frustrating and we have many concerns, issues, and problems with generative AI for lots of good reasons. Yet if we come in with all of that—again, we’re going to lose the students we’re most likely trying to help understand the problems and limitations of these tools.
I’d not only start with a neutral acknowledgment; I’d move into a space that demonstrates a reasonable balance of curiosity and concern. It’s exciting and also opens up questions that I don’t have the answer yet. It can help me do some work such as help write guidelines for activities and yet, it can also do things I’m not sure should be done such as feed my students’ work into AI to evaluate and create feedback. It can help me craft more effective language for the syllabus and also, it can create responses to emails from students that have shared deeply personal things. Some of those feel exciting; some of those feel icky.
Opening up with that balance will help you and them to have a more honest discussion about generative AI and its place in your course. I think it’s important to show the balance—coming to strongly one on either side better demonstrates your own critical thinking about a technology that the students will already have a variety of opinions about. The balanced view shows that you’re taking this technology seriously in way that doesn’t dismiss their own views.
Demonstrate Your Curiosity & Concerns
Ok—so you’ve landed the opening gambit, then what? Here is where I might lose folks and I understand; not everyone is going to feel this comfortable with this technology. But I do think it’s worth digging into explaining what you’re curious about these tools and their outputs. In particular, I think it’s worth sharing the ways that you find it useful for your own life.
This is something I encourage everyone to do, but particularly educators. Regardless of what you think of it, play with it until you find something useful or interesting about it. Don’t just dismiss it. Find at least one thing you would use it for that you think is helpful, unique, playful, etc. I think if one can do that, then having a deep (nondismissive) understanding of why others use it will not just increase significantly, but will also help them find other ways they could use it. And the end goal isn’t just making it useful to that educator. It’s more to build a working practice of finding ways to understand generative AI that can further enhance their approaches and means of allowing it in their course for their students.
It’s also where demonstrating that use to students can improve the discussion. If you introduce how you have used it to solve a problem, why you have used it, and the parameters around that use, you can demonstrate intentional and meaningful use—the same approach you want students to use if they choose to use it.
But it doesn’t have to stop there. I might then shift the conversation into considering all the things we don’t know or don’t know well about generative AI. This is the start of guiding their AI critical literacy skills. This becomes an opportunity to ask questions that you want to linger in their minds.
We don’t really understand how accurate a given answer is unless we already know a good amount about the subject. What does that mean for what we do with its outputs?
We don’t know how the company running the AI accounts for or introduces its own sets of biases in the outputs. What does that mean for how quickly we believe what it says, especially if it confirms our own biases?
We do know that much of the data was taken without authors’ or artists’ permission. We also know that many of these companies used exploited labor to moderate the content and outputs. Meanwhile, the environmental costs of usage are quite significant. What is our role and responsibility in using these harmful technologies?
We don’t quite know the legal status of the outputs and who actually owns them. What does it mean to build with a tool that you may not actually own the outputs?
These are some of the questions you can ask in that conversation to help them think more about what this tool is and isn’t.
Provide Use Cases You Endorse
The next pivot is to identify good uses of the tool. Again, this might be a no-go for folks, but for me, I just can’t imagine the idea that there isn’t ways this can be used in different clases—regardless of whether I would use it or not, students are. I want them to find useful ways that enhance their learning. It’s like comprehensive sexuality education; you can’t play the abstinence game, you have to acknowledge the full range of possibilities—problematic and beneficial and frame it in a way that empowers them to better decision-making.
Even within your class, you can find ways that it can be helpful. You can show them how they can turn it into an interviewing machine to help them figure things out. This is a great tactic when talking about reflection. You can read more about it in this post.
You can provide prompt ideas around certain things that you think make sense such turning it into an editor, an explainer of ideas or hard passages in the learning materials, or a quizzer/interviewer to help assess their own learning and understanding. You can identify AI tools that might be helpful in their research such as Research Rabbit. There are many use cases that can benefit learning—lean into that and show them the possibilities.
You don’t have to highlight all of them but have a few one hand or discuss it in general. Then, you would want to make sure you have references to usages as related to different activities and assignments. The goal is to give them recommendations and representations of proper use.
Open up the Conversation
As I said, this is “the talk” but good talks are dialogue. Create space to actually engage with students, and elicit their own thinking, their own concerns, and the ways they believe generative AI should be used.
To me, this is one of the most invaluable aspects of the conversation because if you frame it well, you’ll actually learn new things—new use cases, new concerns you never thought about, and new ways of seeing your students critical thinking or apathy (which I don’t say as a critique; I think students in a modern higher ed institution in the current state of world affairs, apathy is a fair response).
Ask them about their own views and opinions. Here is a list I created (in part with help from ChatGPT) that you might also ask or pick from:
Have you used any AI tools before? What was your experience and did they met your expectations?
What do you think are the most effective ways to use AI tools for learning based on your past experiences?
Can you identify any challenges or limitations you encounter when using AI in your learning?
In what ways do you think AI can be misused in educational contexts? Why is that misuse? What is lost?
How do you feel about using AI to assist with your assignments and projects in this course?
What ethical considerations do you think are important when using AI tools for educational purposes?
Based on your experience, how do you suggest we ensure that AI is used responsibly by students in this course?
How could AI tools help you achieve the learning outcomes of this course? Are there specific tasks where AI could be particularly helpful?
What skills do you think you might need to develop to use AI tools effectively and critically in this course?
Have you encountered any biases in AI tools you've used? How did that affect your view of the tool or how you used it?
What measures do you think should be implemented to prevent cheating or misuse of AI in academic settings? Is that even the right question to ask?
What do you think it is appropriate for instructors to use generative AI in teaching and evaluating students? What's inappropriate?
What are your thoughts on the fairness of using AI for grading or feedback in courses?
Obviously, there isn’t time to ask all these questions but there might be ways of getting it at it by having smaller group discussions with each group tackling 1-3 questions for a share back in class or as a shared document.
Even if you don’t use these questions, work to engage with them and learn where they stand—some may not surprise you and yet, others certainly will. This candid conversation can also help to build trust and community as you go forward.
Emphasis the Challenge of Deep Learning
I used to do this activity in courses to draw out some differences in our understanding about learning. This played upon the myth of learning styles as well as helping students understanding that friction happens in learning because learning is a hard thing to do. At the end of the day, learning is changing our minds; and that’s not an easy thing to do. If it were, the self-help genre would be nonexistent rather than a bajillion dollar industry. It can be made easier—for sure! Often through activating prior-knowledge, piquing curiosity, and creating supportive contexts and opportunities to re-try along with rich and actionable feedback among other practices. Still, it can be hard for learning to stick.
I would often ask the students how they think they learn best. Nearly all would say they are visual learners. I would have students discuss a passage—about a page long of hard text. I’d have them sit with it and struggle through it. Then, we’d move into discussing it. It would be a challenge and yet, they would be able to engage with it differently and find new things about the passage. A week later, I would ask them questions about it that demonstrated their learning and their ability to apply their learning. At another point in the semester, I would have them watch a documentary. Our discussion of the documentary was always more superficial and limited. They could talk about enjoying it but not digging into it as critically with the text. They often needed to be led to challenge the director’s decisions or even consider the role of the editor. They almost never picked up on the impact of soundtrack and B-roll footage. Similarly, a week later, their ability to apply the learning from the documentary and conversation was lackluster.
Ok—long anecdote but hopeful folks can see where I’m going with this. We have to find ways to help students to understand that ease of digesting information is not the same as learning. Learning requires friction. Ideally, educators are finding the ways to reduce unnecessary friction (of which there is sooooo much in education and nearly every classroom) so that time and energy can be focused on the real friction of learning important aspects of the course.
So this requires unpacking with students a bit around learning, friction, metacognition, and the like and helping them understand that AI can be a tool that services their learning or it can be a tool that supplants it. They may choose to supplant it—for whatever reason, but we as educators can’t really change that. The goal of this conversation is to help them understand
Final Thoughts
The truth is—you will always have students that use it; some of that usage, you will never know about. Students are highly resourceful and there are lots of ways to work around any set of practices we think are foolproof. The goal is not to eliminate it—nothing I’m advising in this series will do that. Rather, I’m a big fan of harm-reduction. I want to create conditions that decrease the fear, angst, and stress that lead students to it, knowing there are still going to be folks that end up using it and the true test of my ability to facilitate learning and growth comes from how I react or respond to those situations.
The totality of this series is to just help use think more critically about what it means to enter into these conversations. We all discuss a desire for students to be critical thinkers; but we too have to be critical thinkers in how we navigate generative AI. I know some folks think it’s a nothingburger or that it is just a fad. I can understand seeing it that way. I can also see that regardless, these practices discussed in this series are still really important because they rely on relationship-building with students and building rapport that will better impact learning regardless of whether we’re still talking about generative AI in 10 years or some other technology.
What am I missing? What would you add? What feels challenging for you to grapple with?
AI+Edu=Simplified by Lance Eaton is licensed under Attribution-ShareAlike 4.0 International
