Ethical AI for Instructional Designers With Dr. Dwayne Wood

Welcome & Series Wrap-Up

Jackie Pelegrin 0:00

Hello and welcome to the Designing with Love Podcast. I am your host, Jackie Pellegrin, where my goal is to bring you information, tips, and tricks as an instructional designer. Hello, instructional designers and educators. Welcome to episode 158 of the Designing with Love Podcast. I'm excited to welcome Dr. Dwayne Wood back for part five, the final conversation in our mini-series. Across the series, we have explored generative AI's impact on academic curriculum, multimodal strategies for adult learners, adult learning principles, and how AI can support innovative instructional design. Today we're closing the series by looking at the evolution of AI and its ethical implications in content creation. This is such an important topic because as AI tools become more powerful and more accessible, instructional designers and educators need to think carefully about accuracy, bias, transparency, authorship, privacy, and the human responsibility behind AI-generated content. And I we've discussed this a little bit, uh Duane, but I'm excited to delve deeper into this. So welcome back to the show.

SPEAKER_00 1:11

Well, thank you. Thank you.

Jackie Pelegrin 1:13

Appreciate it. And what does that evolution mean for educators and instructional designers today?

SPEAKER_00 1:23

Yeah, and you know, it's but the the the evolution is so quick. It's so fast, right? You know, um just a couple years ago, it we had the first kind of you know open source chatbot where we're, you know, it works with with with text. And now today we have full authoring tools where you can upload some text and it'll produce a um e-learning based upon what you what you inputted. Um and they've gone from text to multimodal, right? We're we're producing images and video and um all you know, the whole gambit of media that an instructional designer works with. Um so it it's hard to say what it's gonna look like, you know, just in a couple weeks with some of the advances that we're seeing here. But I I think there's an important um piece to think about this is that it's not just the you know, the idea that it's it simply creates more content. Um it's creating more content quicker in more formats. And in a lot of times, it's you're creating a media that generally would take maybe some specialized um technical expertise that now anybody can do. Right. So um that's where I mean it's gonna continue, I think, to get um you know more powerful and more powerful. Um and I think I think we're gonna talk probably later, but you know, but that doesn't mean we it produces exactly what we're looking for and what needs to be used, right? There's there's there's a role um for the instructional designer in that review process and and putting all that together and understanding how these technologies work to produce that that um learning event you're trying to put together.

Jackie Pelegrin 2:58

Right. Absolutely. So true. So, Duane, as AI becomes more common in content creation, it is important to think carefully about the responsibilities that come along with using it. We've talked about this in our previous conversations. So, what ethical concern should

How Fast AI Content Is Evolving

Jackie Pelegrin 3:18

instructional designers pay attention to when using AI to create course content, learning materials, assessments, or learner-facing resources?

SPEAKER_00 3:28

No, and I think this is like the big question. Um and I think there's still a lot of like unresolved kind of issues in this. And I boil it down to like four topic areas. Uh, one is accuracy, right? We already know that these these tools will make things up, right? Or um produce something

Four Ethical Risks To Watch

SPEAKER_00 3:44

that sounds great, right? It sounds, but you know, it's not accurate. Um, you know, the other one I think is privacy. When we talk about you know what data is going into these systems and where is it going, right? I think that that that's a concern there. I think we also need to think about intellectual property, right, with copyright, you know, um, what's the lines on fair use and and thinking through some of those those pieces there. And then lastly, we need to think about accountability. So whatever these systems produce, who's accountable, right? Um and you know, just thinking about when we talk about the um accuracy, right? These things, definitely incorrect statements, right? Invented sources. I see with students all the time, if they try to, you know, um, I'll I'll look at a source to see if it actually exists, and you know, it's made up, right? But it looks, it looks good. It'll even produce a DOI, right? Um, but it it it's you know, um distortions, right? It you you might put a document that you want to get a summary, and you might get a little bit of distortion. Uh, and maybe some of the nuances of the argument, whatever's in there, doesn't come through in the summary, right? So the stuff we have to be kind of aware of. Um when we talk about privacy, I almost I also tack on kind of privacy and and and fairness when we talk about bias, right? Um, these these tools can can um actually have bias within their training database, but also just the way you ask the question or the way your prompt is oriented may inject bias. Actually, I would say that in most of these systems you use, you're automatically gonna get confirmation bias because um these are the ultimate yes person, right? They're gonna give you what you're looking for. So we have to be careful of that and just understand that. Um, but even in the we've seen these trend in the database, I've seen a bunch of articles and stuff and come out where you know they ask for an image of something and they get the stereotypical white male, right? Um so it's already kind of injecting some level of bias. So we just have to be aware of that. It doesn't mean we can't use the tool. We just need to be aware of that as we're we're we're using it. And I and I think when we talk about data and the privacy, you know, we have to be really careful what we're putting into it. Not only just on privacy of data, but that this is on adds to that intellectual property or copyright aspect. So if you take something that's copyrighted and you add it to a prompt that you're gonna use, well, that that's probably maybe going into the training database. Now have you exceeded the fair use because now it's distribution somewhere, right? Um, so we have to be really careful if we're if we're gonna use a source, let's make sure that it's got a Creative Common license that allows us to do that, or that it's you know, it's public domain, right? So that you can go in there and and use it. So there's some things we have to kind of really, really um think through to make sure we're gonna be operating within copyright. And and again, I think they're still trying to think through this. Um, there's been lawsuits, right, back and forth uh uh about this. And ultimately, whoever the user is, and in this case, we're talking the instructional designer, needs to take accountability for whatever's produced. It's not the system, it's it's the user.

Jackie Pelegrin 6:54

Right. Exactly. Um, yeah, being accountable for what we're inputting into the tool, that's important, Dwayne, because I think a lot of instructional designers may not think of that. Like what they're putting in to the tool may already have a bias to it, and we may not realize that that's true. Um, that's something I I even need to remember myself when I'm prompting the tool, right? And what kind of output I get is it's kind of like, you know, what you put in is what you're gonna get out of it, right?

SPEAKER_00 7:22

Yeah, absolutely. And then even on the student side, right? If if they go into the library and pull down some articles and they're copyrighted and they put them into a system, are we now, is the school now liable for a copyright infringement?

Jackie Pelegrin 7:35

Right.

SPEAKER_00 7:36

I I I I again they're they're still kind of sorting this out. Um, so um I I just think we have to be we just have to be aware. And as long as we're aware, we're able to make better decisions when we're interacting with these to with these tools.

Jackie Pelegrin 7:50

That's so true. Right. Yeah, because every institution, you know, they yeah, they have to be careful about intellectual property with textbooks and with outside sources. Um, you know, with the institution I work for, when we do curriculum development, I think sometimes you're right. Students and faculty even sometimes don't realize it because they'll go and share an article or a resource in the class and they can do that, but we just can't put it in the course shell. And um, yeah, and so they they forget that. And they're like, oh, can you just put this PDF in the course? And it's like, no, we we can't do that. That's uh copyright infringement. We can point to the website where that PDF is, but but we can't do uh just put a PDF in the course, yeah. Unless it's produced by the university, then then we can do it.

SPEAKER_00 8:37

But yeah, and and you know, that's that that's that distinction in fair use, right? Um if if if we're if we're using a part of it or you know, and we got there's some rules in there we can work with, but if we're just taking the whole and giving it to somebody that that now it's theirs, now you've gone from fair use to distribution, right? So now you're in now we're potentially in in I mean it it's not a cut and dry. There's like layers of fair use like to make sure you're you're you're doing it. But um we just gotta be you just gotta be careful, right? And thinking through that. You know, when I work with um in building like in higher education, I always look for you know OER or open educational resources, um, or you know, Creative Commons, right? Because then those things you can you have a little more, you can you can work with them more. Um, you know, if if you're not familiar with Creative Commons, I would tell you go out there and just do a search Creative Commons and you can see the different levels that things are um marked, and then it gives you like restrictions uh based upon that level of license. Um it's a very good thing for instructional designers to kind of know about it.

Jackie Pelegrin 10:15

Right. So in the in the essence of AI, then instructional designers should be careful not to just take something from the internet, right, and put it into an AI generated tool, because then that can that can um cause them to get into a sticky situation, right?

SPEAKER_00 10:33

Absolutely, right. Um, uh I I think the image generators have gotten better. Um, but I know when they first was doing them, there was some times that I would ask it to do an image and I would actually get watermarks from where it took an image from shutter stock, right? Um when it was putting stuff together. So again, we have to be we have to be very careful. I the continue to get better, and the technology is going to continue to get better, and you know, there's a lot of smart people thinking about the same things we're talking about here, and they're they're figuring out how they can work and make this um, but there's still a lot of unanswered questions. We're kind of in that Wild West um, you know, realm right now, and we just have to be careful and and you know, make sure we're thinking through that and that we don't put ourselves and our organization we work for maybe in in some kind of um situation.

Jackie Pelegrin 11:20

Right. That's so true. You know, I often wonder sometimes when it's when an AI tool is generating

Copyright And Course Content Reality

Jackie Pelegrin 11:27

images and then it has a a person in that or people, you I often wonder, is it taking the image and likeness of someone that is actually in existence out there? You know, and I'm like, I hope not. That's why I kind of prefer not to have people or um anything like that in in my images. I mean, sometimes I do want that. I want to show collaboration or something like that, and I think it's good to have that. But I try to, I don't know about you, but I try to kind of um steer clear of that as much as I can, unless I really need it to be incorporated to represent something that I'm trying to come across with my message or something. Um do you kind of think that too is a good idea?

SPEAKER_00 12:07

I do the exact same thing. If if if having a person in the image is not needed for the you know the purpose of that image or that w whatever the media piece, I don't use it. Um I you know, I I'll I'll stay with more of the you know the the imagery or abstractness um because you're right. There's just a lot of pitfalls. There could be a lot of pitfalls, right? And if if you don't have to use it for the purpose you're doing in, you know, don't don't do it, right?

Jackie Pelegrin 12:33

You can you can do it other means you know, and it's interesting too, because I did have one generated for my one of my companion blog posts for an episode. Uh and I noticed that when it generated the image, it had uh it had a woman in it, and it was uh it was representing like an instructional designer working through something. But I noticed that they had it had her holding a pen, but the pen did not go all the way, it went like partly through her between her fingers, but it didn't go through on the other side. And I looked at it closely and I'm like, that doesn't seem right. I'm like, so I decided to just have the person taken out and and I was like, but it's still I I'm noticing that it still does weird things that I'm like, that doesn't look right. Yeah, like that's odd. So yeah, you have to do look out for those things.

SPEAKER_00 13:20

Yeah, I think it has to do with like how it's trained. Um, you know, you know, there's different color contrasts and shadows and all of this stuff, right? It's it's it's we're they're still making that better, right? Uh and they're drastically better than they used to be when they first started coming out, right? You would get have you know three legs or you know, the you know, arm where the leg should be, and yeah, you know, all that crazy stuff. They've gotten better. Yeah. Uh but they're still like you pick up on those little nuances.

Jackie Pelegrin 13:44

Yeah, exactly. Right. So important. Or it misspells. Oh my goodness. I still notice that it misspells words. Like it's uh like half Greek and half something else. And I'm like, what? Oh my goodness. Yeah. But it's yeah, as you said, it's doing better with that too. But you really have to be aware as an ID, right? And catch those types of things um and not just take the first output all the time that it's giving you. So important.

SPEAKER_00 14:08

You know, and honestly, when I when I use the image, I don't actually use the image that it produces, I use the image that it produces so I can build it. Um again, I I could probably use like in some of the images I do are pretty basic. So I mean I could probably use what it creates. Um, but I like to do that because it's easier for me to edit in the future because I have it built. Right. And I have it. It's in my library, right? And I don't need any special editing tools. Um, because again, most of my clients I work with, my authoring tool is PowerPoint. So I'm building the image in PowerPoint, but now it's editable, right? I can I can edit it and so I use it more as like an idea generation of what it could look like or the design aspects of it. And then I kind of create that in my for my own. Again, I'm not really saving time. Well, I guess I'm saving a little bit of time because of the idea generation brainstorming pieces. I'm I'm making that happen quick. Uh, but the building, I'm not saving time because I'm still building it. Um but I do that for some, you know, again, more sustainable reasons, because I'm probably going to edit it in the future. So I want to have it as part of my course document got to documentation where I can edit it for any future updates.

Jackie Pelegrin 15:19

All right. That's true. Yes. That goes into kind of the next part of our conversation and the question too, because you're touching on this right now. So there's also that practical challenge of balancing efficiency with quality. You're kind of kind of hinting on that a little bit. So, how can educators and designers balance the efficiency of AI-generated content with that need for accuracy, originality, inclusivity, and even human review as well along the way?

SPEAKER_00 15:48

No, and this is a I really like this question. Um, and I had to put some additional thought into it because I have my own internal process, but I never like sat down and said, uh, hey, this is how I think I would want somebody to approach this. So I'm gonna, I'm gonna uh kind of talk to this from my own process, the way I do this. And what I would say is that, you know, one of the things that these can do is you can generate first drafts very quickly, right? So that's an advantage, right? It helps you build out those pieces. Um, but as you mentioned, there's a bunch of things you would need to do to make sure that that's good, right? It's accurate, you know, you're you know, all these other pieces in there. Um, so I almost think of this as almost like risk levels. So what am I doing with the tool? If it's low risk, I might have a little low less of a review process because if I have a full review process, I'm not really gaining much in the efficiency, right? I'm using the same amount of time to get this about the same product. If I was getting a much better product, then maybe I would give up quality for time, right? But having that kind of um cost benefit analysis, right, as as as you're as you're thinking about it. If it's a a little higher risk, I definitely want to have like a no kidding, like full, make sure I'm uh double checking all that good stuff and doing all this. But if it's pretty low risk, maybe I just have um, you know, some and what does it mean to be low risk or medium risk or high risk? It's just gonna take experience, honestly. I think. Um I I I was wrecking my brain, I was like, how do I explain that? How do I explain that explicitly to someone? And I think it comes down to their context, um, their knowledge and comfort level with the content they're

Image Generation Pitfalls & Workarounds

SPEAKER_00 17:27

working with, right? So I mean, there's just a bunch of things I I think they're they're they're working in. So like something that um to me would be low risk is maybe I'm gonna I have a I have some content that I'm pretty familiar with, and I have it generate 10 multiple choice questions with the answers. That's pretty low risk. I'm still gonna review it anyways, right? But I might not do a full um piece where maybe I'm gonna have some content that I wanted to help generate the text for a student-facing document or aspect, and it's content I'm not as familiar with, that's a little higher risk. So I'm gonna make sure I do the full review process. And again, I still may save time there because it would take me a long time probably to do that with content I'm not familiar with, or I would need to work with a SME that you know, there's coordination time, all that type of stuff. Um, so uh one of the things I think of when we talk about working with subject matter experts in this is could I get the content close so then I'm having a subject matter expert review it? So now I've kind of sped that process up a little bit, and I still have the human review process. So I'm actually gaining some efficiencies there. So with all that said, is I think you have to really we have to really think about our context, the risk level, and what we're working with and where we're going and what we're doing and do that assessment to ensure we get efficiencies. Because if I have a simple task and I do it, and then I do a full review on it and I and I'm I'm digging into it, I probably didn't save any time. Actually, it's probably taking me longer. Um, you know, same with if I if I get into that chain of thought prompting trap, right? Um you put something in there and you go into like, that's not exactly what I want. Okay, you give it something else, then it says, oh, well, I want to adjust this, and you do that again, and then before you know it, it's 60 minutes later, and you're still not to a product you like, right? Um it you didn't really save time by using that tool when you could have just done it the normal way, you know, non-AI and probably been done in 30 minutes. Right. So it's coming to understand what you can accomplish, you know, and and the the quality of the outcome and then the risk and working through that to to know where those efficiencies you can gain. And unfortunately, my answer is it depends. Um, you kind of have to know and and and just work through your processes to know your processes and know your what you're working on and and make those decisions. Um but absolutely absolutely I would never say just create something, copy and paste it and put it into a product. To me, that's like a that's just no, right? Um there's gotta be some level of of review on it. But that might be less if it's a low risk type environment.

Jackie Pelegrin 20:03

Right. So true. Yeah, that's a great, great point, Dwayne. I love that. And it kind of you know brings us back to our last conversation a little bit too with the technology, right? We shouldn't we shouldn't use technology or we shouldn't use AI for the sake of it, should be used with purpose. And yeah, absolutely. I love that. So another important part of ethical AI use is being clear about how and when AI is involved, which is something that we're you talked a little bit about uh previously just now. So, what role should transparency play when AI is used in the content creation process?

SPEAKER_00 20:40

So I think it's absolutely essential. Uh there's gotta be some level of attribution. Uh now, does it need to be uh, you know, um a full attribution every single time? I don't think so. Um but there still needs like give you an example. If if I use an AI tool to check my grammar, I I I I I probably would still put, you know, I AI was used to check grammar or something. So someone that kind of knows, that's different than if the entire thing was created by an AI tool, right? I need it, I need to have a full attribution statement there. Um because when we think about this in kind of the bigger picture, if I send you a document for review and it has no attribution statement on it, and you take it as my work, and you you may take it because you have a level of trust in my work. Um, but if I put an attribution statement that I used AI on it, you might look a little deeper because now you're looking at the tool being used and not the person that produced it, right? You know, so I I just think it's absolutely we have to be completely transparent. Again, who owns accountability? It's the person that put it together, the user. Um but we need to attribute that we use some of these tools so that we don't do some of those, you know, screening. We when we do the screening and review process, we know to look a little maybe a little bit harder or some of those specific things we know AI doesn't do real well, right? So that we can we can actually focus in on those things and find them during the review process um so they don't survive, right? That review process as we're going through. So I think there has to be some level of attribution. And but I think it varies depending on you know the use that it's being done. Um, you know, and again, maybe in the final product, like the student facing product, there's not an attribution statement. But during that development process, there is as it goes through the cycles. So everybody kind of knows that that that a specific tool in a specific way AI was used to help them review the pro the product as it's going through the process. So that that that again, not and again, I think. I think it's even on the student side, right? When you're submitting products, there should be complete transparency there. You know, what's allowed with AI? You know, how did I use it? Um for a couple reasons there. One, for the reasons we stated, but it's also a learning event.

unknown 22:51

Right?

SPEAKER_00 22:51

Because if you if you articulate specifically how you used it, right now it's it's it's concrete. It's explicit, it's explicit. So I think that's that's an important step when we're when we're learning to work with it, right? So even I guess even as a structural designer doing that, right? You're you're now you're making explicit how did I use this tool producing it in the production of this product. You make it explicit when you when you attribute it.

Jackie Pelegrin 23:14

I yeah, and I even think that even in the like you said, in the development process, even working with subject matter experts, I think that they should even attribute it, right? And I don't think that's clear up at up to this point because I work on the curriculum development side and the faculty don't tell us whether they used AI or not, but you can kind of see some of the the telltale signs of it that it was used. And so yeah, I'm wondering if during that process, if the subject matter expert should uh disclose that too, um, because they're expecting

Balancing Speed With Quality By Risk

Jackie Pelegrin 23:49

students to do it, right, and their work. So why why can't they do it? Um it's just modeling good behavior overall, right?

SPEAKER_00 23:57

Well, on top of that, it should be in compliance with whatever organizational policy is in place.

unknown 24:02

Right.

SPEAKER_00 24:02

Right. So if there's an organizational policy that there's a hundred percent transparency, then yes, absolutely. Um, but I'm not aware of a lot of organizations that will have a hundred percent transparency um as part of their policy.

Jackie Pelegrin 24:16

Right.

SPEAKER_00 24:17

Right, you know, um so and and again, we're in the Wild West here, right? I mean, we can use academic journals as an example here. Some academic journals say no. You know, if you s if you want to submit an article to their journal, you have to um, you know, signs saying that you have not used AI at all, right? And then there's other journals that allow it, you know, part of the review process, and because they're they're banking on the idea that the author is the one that's accountable, right? So there's not a consensus there. So I I think it's the same thing in our field. We're seeing the same thing. There's not a consensus. Um, but I I would always err on the side of making it transparent. You know, hey, I used AI to do this, you know, as part of this product. And again, it doesn't have to be in the final product. That that attribution can be removed, right? But during the development process and review process and the quality assurance process, you know, it's attributed.

Jackie Pelegrin 25:17

So important. Absolutely. Attribution. Yeah, that's a great reminder. And I think it brings to the forefront why it's important to do that. Um, because we don't want uh to get in a sticky situation. I don't think we don't want our company or organization to get into that situation either. So yeah, very important. So doing onto our bonus question, what is one ethical red flag that should make an instructional designer pause before using or publishing AI generated content?

SPEAKER_00 25:48

So uh again, some really great questions in this, in, in, in this. So uh so you have to bear with me on this one. If you produce something and a an accountable human, so a person that's part of the process, can't explain, verify, or defend that content, a red flag should go up. Right? Uh and I and I and I I did this kind of, I flipped it the other way because I I when I work with students, a lot of times to make uh an assessment a little more AI resistant, you have them present, right? Because then they're they're they're they have to defend something that they that they did. And if they used AI, you can tell pretty quickly that they don't have the depth of experience because they didn't go through the cognitive effort to produce that product. So you can tell right away. So that's exactly if you if you can get that response where no one can go, oh, I can explain that, or I can verify it, or I can defend it, um, you you you probably have some problems because you can't you you you don't have the human in the loop anymore, right? You you let the you let the tool just produce something that no one can defend or verify or explain. So you might have a problem there. That's when I would say, you know, you you the the red flag should pop up and be like, okay, we need to look at this.

Jackie Pelegrin 27:04

Right. And and pause and not move forward until you do that, right?

SPEAKER_00 27:09

Absolutely. Yes.

Jackie Pelegrin 27:10

Yeah. Right. So important to just take a pause for a moment. Absolutely. So, Duane, as we close out this mini-series, I can't believe how fast this went, but um I'd love to bring the conversation back to what educators and instructional designers can take forward with them. What final advice would you give to an instructional designer or educator who wants to use AI responsibly while continuing to design meaningful learning experiences?

SPEAKER_00 27:38

Yeah, and uh that that's important, right? Um, because there's a ton of like trainings and stuff you can go to, but until you actually sit down and try to use it, um, you don't, I don't think you get a full appreciation. Uh so my advice is pick us a low-risk use or low risk task. Right? One that you have the domain knowledge to judge and evaluate against whatever the system can produce. And ask it to do a lot of the instructional design tasks that you would normally do. And don't judge what it produces based upon like amount. That's not a good metric there. What's a good metric is what what does it produce that's accurate that allows for an educationally meaningful uh process, right? So it's not just the volume that it produces, because these things can produce uh volumes, right? But what did it produce stuff that works to produce uh a meaningful educational experience and use that as a metrics to judge what it produced? Again, thinking in an instructional design context as we go through this. So really low risk staff. Again, one that you're very familiar with the knowledge, you know it, right? So that you can actually judge what it produces. And I think you'll see if you run that, I think you'll see some of the nuances of these tools. And you'll start to understand some of the limitations and boundaries that these may be producing, especially within our instructional design uh context. And I think that'll help you think through the as you get into you know actually production and using these tools in your instructional design process.

Jackie Pelegrin 29:20

Right. I love that. Yeah. The low stakes start with something low you know that you're familiar with. I like that, because then you're not getting into something risky or an unknown, right? Early on. And then you can build up your expertise and your knowledge in in this area that's like you said, it's the wild, wild west. And we're still navigating it. Yes.

SPEAKER_00 29:43

You know, and and when you do this too, know your organizational policy. Um if you don't know if there's one, ask. Right? And and know what it says and know what it means, right? It's it's one thing to read it, but it's one thing to know what that actually means in practice, right? So that you you know the boundaries of what the policy your organizational policy is, so you're not you're not in violation. Uh, because different organizations have different levels of risk. Some it's open and some they have a little more restrictive. So you want to make sure you understand that organization you're working for's policy on AI use. And they may have some very specific things on attribution, like we talked about before, right? So you'll make sure that you're following those policy that they have laid out.

Jackie Pelegrin 30:31

That's so true. And the company I work for, the organization, they we we're a service provider, so we provide services to institutions. So not only

Transparency & Attribution Standards

Jackie Pelegrin 30:41

do we have to follow our policy as an organization, but we also have to follow the institution's policies as well. So it's a almost a double-edged sword in a way, but it but it's good. I mean, I think it's important. So I think uh it's important that we don't lose sight of that, right? That sometimes you may have to follow multiple policies and hopefully they don't contradict each other. Hopefully they are in sync with each other and follow a lot because then that would make it more confusing, I think, if they contradicted or weren't in support of one another, right?

SPEAKER_00 31:13

Yeah, and and I think you bring up a great point there, right? Um a lot of us are in organizations that, again, like you said, are supporting other organizations. Um, so there's multiple policies to navigate. I think that's a a very important um aspect you brought up. Uh, because you know, you're you're supplying that service for that client, you know, and you got to know their policy. Uh especially if you're designing maybe an educational experience that includes AI.

unknown 31:39

Right.

SPEAKER_00 31:39

You know, so you know that that could be you know, could be a problem, right? I mean, if so knowing those things are today are critical for an instructional designer. You have to know those things. Absolutely. Know those governance procedures.

Jackie Pelegrin 31:53

Right. That's so true because yesterday I was doing a final review on a course revision. It's a psychology course on coaching, and that's an undergraduate course. And one of the assignments, the student needed to put in an AI prompt, and they had to imagine that they are um they're creating a PowerPoint presentation, and it's for something that they would do in real life. And so they had to either take the prompt that was in the instructions in the assignment or something similar, put it into the AI tool. And I noticed some errors in the way it was worded. So I gave some suggestions and it repeated itself in the instructions. As it said, use an A uh Gisu approved AI generated tool. And then it said use the, and I'm like, so okay, so we're saying use twice. So that was like that didn't sound right. But not only that, but at the at the end of that prompt, it said for them to generate three to five peer-reviewed resources using the I AI tool. And I'm like, uh, I don't think it's gonna do a good job of that. It's going to give fake sources. So I was like, I don't know if we should be asking students to do that using AI. So I, you know, I put my concerns in there on that one.

SPEAKER_00 33:05

Yeah, that I I I would fully support that. Um, because you know, another thing it does, you know, that I if if you're asking for sources, is because of the training database, it tends to give older sources. So it doesn't usually give more recent ones. It also likes to recommend books because books, um, there's a book review in it in the training database, and it has a little more statistical probability, right? So it tends to recommend those. Um, which in in our in my classes, under four weeks, if a student puts four books, they're reading it for references. That's not very reasonable in a four-week course that you're gonna read full four books, right? Um what you're doing is just mining from the book review to support some kind of position, which is not what we're trying to achieve here. Um, right. So again, there's things when when you start seeing these, you start recognizing some of these things, and you can you can immediately spot. And again, it's not like I call out the student and you know, and it's it's a it's a pun, it's a teaching moment. But say, let's talk about your sources. How did you get them? Uh, because I find sometimes when we start talking about that, they didn't necessarily know how to navigate the library, or they didn't understand maybe how to use the filter tools in Google Scholar, or how to evaluate a source. So it brought up other things we could talk about and you know to improve for future, right? As they as they work through their academic career. Um, or even like I find all the time the students don't know um the great librarians that work at the school, and all you have to do is pull them on chat. And and and they're fantastic. Um, you know, so sometimes just mentioning that and showing them and hey, looks, let's let's hit the chat button and see what we get, right? Um, and they get to talk to a librarian, you know, an expert and help them navigate the library, you know. So um again, it it's a teaching moment. Uh, but we have to it when you when you're in that instructor or teacher or facilitator role, you have to be able to recognize those things and kind of like, you know, call them out. Again, being transparent, right? If we were completely transparent, it'd be easy to have that conversation. Uh, but now we're trying to figure out when they do it. And we know these AI detection tools are not great, right? Um, you know, so we have to be able to, you know, find those things that we recognize and be able to have that conversation with the student.

Jackie Pelegrin 35:14

That's so true. And and we don't want to accuse them right up front, right? We want to, we uh, yeah, like you said, you want to make it a teaching moment and and just kind of explore it and ask them, you know, where did you get, you know, where did you get this or how did you come across it and and and not make it to where they feel like they're being um reprimanded or in trouble because then you know that that ruins the experience for them moving forward, right? So we want to be careful how we go about doing that. So I like the way you you do

The Red Flag Nobody Can Defend

Jackie Pelegrin 35:43

that with your students. That's great. I love that.

unknown 35:45

Yeah.

SPEAKER_00 35:46

Yeah. And you know, it again, it helps like again, you we're we're working with we're working with people that we want them to be able to develop those skills of research to understand theory and then be able to apply that in practice, right? So there's you know, uh, research skills, critical thinking, all that is is absolutely important to develop. And, you know, sometimes just having that conversation helps develop that or or opens up the, you know, like, oh, I didn't think about it that way. And, you know, and right there's the learning moment, um, which is what all teachers look for, right? You're you're super excited when you see the light bulb go off. It's like, whew, that was the best thing ever.

Jackie Pelegrin 36:22

Right. So true. I love that. Well, Dwayne, thank you again for being part of this five-part mini-series and for sharing your expertise throughout each conversation. I really value that. We've covered so much ground from curriculum and adult learning to innovation, AI-supported design and ethics and content creation. This has been such a valuable series for instructional designers and educators, and I would say even myself too, who want to understand AI without losing sight of the human responsibility at the heart of our work. So, for anyone who wants to keep learning from you beyond the series, what's the best place to follow you or connect?

SPEAKER_00 37:00

Uh please connect with me on LinkedIn. Uh definitely would love to have you know further conversations. Um, or you know, or just my national university email address. Uh again, continue the conversation there.

Jackie Pelegrin 37:12

I love that. Yes, because we're all a big community, right? And we're always learning and growing. Um, they say the moment we stop learning or growing, we should be concerned, right? Yes. No, I love that. Great. Well, thanks again, Dwayne. And and hopefully we'll do another series down the road. Um, as things evolve and continue to change, I'm sure there will be opportunities

Practical Advice & Policy Navigation

Jackie Pelegrin 37:34

down the road for us to do another mini-series down the road.

SPEAKER_00 37:38

I would definitely look forward to that. Thank you.

Jackie Pelegrin 37:41

Great. Well, thank you. We'll stay connected and take it from there. Thank you for taking some time to listen to this podcast episode today. Your support means the world to me. If you'd like to help keep the podcast going, you can share it with a friend or colleague, leave a heartfelt review, or offer a monetary contribution. Every act of support, big or small, makes a difference, and I'm truly thankful for you.