Bonni Stachowiak [00:00:00]:

Today on episode number 517 of the Teaching in Higher Ed podcast, thinking with and about AI with Eddie Watson.Production Credit: Produced by Innovate Learning, maximizing human potential.

Bonni Stachowiak [00:00:20]:

Welcome to this episode of Teaching in Higher Ed. I’m Bonni Stachowiak, and this is the space where we explore the art and science of being more effective at facilitating learning. We also share ways to improve our productivity approaches, so we can have more peace in our lives and be even more present for our students. 

C Edward Watson is the Associate Vice President for Curricular and Pedagogical Innovation and Executive Director of Open Educational Resources and Digital Innovation at the American Association of Colleges and Universities, AAC and U. Prior to joining AAC and U, Dr. Watson was the Director of the Center For Teaching and Learning at the University of Georgia, where he led university efforts associated with faculty development, TA development, learning technologies, and the scholarship of teaching and learning. He continues to serve as a fellow in the Louise McBee Institute of Higher Education at UGA and recently stepped down after more than a decade as the executive editor of the International Journal of Teaching and Learning in Higher Education. His most recent book, which he’s here to talk with me about today, Teaching with AI: A Practical Guide to a New Era of Human Learning, is coauthored with Jose Bowen, who I’ll be speaking with on the next episode. Dr. Watson has been quoted in the New York Times, Chronicle of Higher Education, Campus Technology, EdSurge, Consumer Reports, UK Financial Times, and University Business Magazine, and by the AP, CNN, and NPR regarding current teaching and learning issues and trends in higher education.

Bonni Stachowiak [00:02:18]:

I’m so glad to be welcoming back to the show Eddie Watson. Eddie Watson, welcome back to Teaching in Higher Ed.

Eddie Watson [00:02:26]:

Thank you for having me again. I’m really excited to have this conversation with you.

Bonni Stachowiak [00:02:29]:

I have been looking forward to speaking with you again, and it has been many years. I I had sort of lost track of time. One of the things I think back fondly about our conversations, both for the podcast, but we also have been able to share a meal together and meet in person many years ago, is just your commitment or at least how you come across to me is really being rooted in research and evidence while also I don’t I feel like an open mind is too is an inelegant way of describing you, but just just you seem like a person who’s okay, that it’s not all figured out yet and that almost enhancing your curious. Am I am I right about that? Are you okay with everything not being all figured out and tied up and that just sort of driving your interest in research areas?

Eddie Watson [00:03:17]:

Dave gotta say that that is a very apt description. And, you know, given how we don’t see each other every week, I’m surprised at how much on target what you just described, I think think is for me.

Bonni Stachowiak [00:03:29]:

Yeah. I’m working on right now curating and writing a open textbook. And so midway through my class, I asked the students for some feedback on it, and it was funny because one of them I I had different features of the open textbook. It’s it’s got the ability to search which many learning management systems you can’t do a search across all of the even a course, let alone the all of the courses someone might be enrolled in and things like that. And so one of them had commented just about how meaningful it was to them to have access to this textbook without any expense on their part from day 1, and I thought, Eddie’s smiling even though he doesn’t know that. I mean, I really learned that from you is just and it has been a bear to do. I’m not done as of you and I talking Dave, but that that wrestling and how hard it is to do something like that, you really have inspired me without probably even realizing that all these years later from your research around open textbooks and the way that act so significantly contributes to student success. So I just wanna thank you for your continued research.

Bonni Stachowiak [00:04:30]:

Even though we don’t get to talk very much, I continue to learn from you and draw from the inspiration you provided in past conversations.

Eddie Watson [00:04:36]:

Well, thank you for the kind words. I feel like a lot of what I do and I think a lot of us do in the classroom or CTL directors do. We we are sometimes even when we have students in front of us, we’re still sort of a step or 2 disconnected from the student experience and then to have these kind of conversations with students and go, oh, yeah. That’s right. That’s that’s exactly what I was hoping whenever I decided to dedicate this time that this would be the impact. But, yeah, we don’t we don’t often recognize the real impact that we’re having on our colleagues, our students, our family, unless we have these kind, gentle moments like you just described.

Bonni Stachowiak [00:05:08]:

Yeah. And, of course, the open education movement is one that continues on today and that we are finding even more information about today. And today’s conversation is gonna be about a book that you coauthored with Jose Bowen about teaching with AI. And it is another area where I feel this need for us to become better at navigating ambiguity. To what extent did you discover in this process of writing and collaborating and working with so many faculty that we need to embrace kind of this sense of being okay with ambiguity that we don’t have it all figured out today.

Eddie Watson [00:05:46]:

I think I think I began to recognize that whenever my sons, who are just college age, one just graduated from college, when they showed me chat GPT right after Thanksgiving in 2022, and as they were showing it to me, they described it to me, and I was like, yeah, right. But then as they were showing it to me, my son, Liam, very much remembers me saying, oh, no. So I was beginning to get a sense of, like, oh, this is what the ramifications are just in terms of being able to produce writing. Of course, it does so much more than that. But really since then, there’s been this space of ambiguity, and there continue to be challenges where there aren’t clear answers. It’s not like, oh, well, then just do step 3, and then that’s gonna solve this issue. It’s there’s a lot of gray area in solutions, and a lot of it has to do with our own sense of of ethics regarding technology and how we approach our students from, really, from assignments all the way through academic integrity. So I think I’m always comfortable with ambiguity, but this is certainly a space that provides lots of opportunity for ambiguity, and it is like an onion.

Eddie Watson [00:07:00]:

It’s like, oh, wait. There’s this, and then there’s this also aspect, and then also what’s happening within the world of work, and how does that influence what we’re doing in our classrooms. So I guess I’m fortunate that I’ve I’ve often been comfortable with ambiguity because I feel like this is something that’s created a lot of cognitive dissonance for higher education, really kind of at every level.

Bonni Stachowiak [00:07:21]:

Yeah. I I wanna I wanna first ask you more about the oh, no. Can you go back in time and think about our brains are fascinating to me that all the different ways that, you know, a single thought can kinda burst into this cloud of different thoughts. What were some of the things that were coming to mind for you at that time?

Eddie Watson [00:07:39]:

The the asking it to write a paper of a certain length with a certain number of words and that it produces it and that the quality is reasonable and fairly indistinguishable from the voice of a human.

Bonni Stachowiak [00:07:56]:

I definitely relate to that, the oh, no, and and and you mentioned about being able to write a paper of a specific word count. I can recall early. I’ve been I’ve been teaching in a higher education context for about 20 years now, and I remember seeing colleagues that would get out of RULER, and they would measure the the hard copy papers that would be submitted, and and make sure it was exactly whatever the style of writing required. And and please, listeners, understand that I I recognize the importance of us being able to adapt our writing style and our formatting to meet whatever the specifications are. But I can just recall thinking, I do not have that kind of time to pull out a ruler. I mean, I can’t even find a measuring tape when I most desperately need one, let alone I wasn’t back then even having students turn papers in that was hard copy. So I just thought I had this all figured out, and then it I was gonna switch to it being a word count, you know, that versus that I was gonna sit there and measure things. And, of course, that was a a huge oh, no for me.

Bonni Stachowiak [00:09:02]:

And another one that I’m running into more recently, Eddie, is not just the word count, but another way that I would attempt to assess how students were thinking about what they were taking in was to ask them to explain a concept to me as if they were talking to an 8 year old. And Dave I used to find that was a really good gauge. And I think for students, once I could get them to trust me that this is I’m I’m trying to help us all learn together. I’m not trying to trick you or whatever, but if we think in terms of childlike ways, we can see, oh, wow, I don’t understand this. And and you mentioned having kids. I I also have kids, and our our our kids are now older. I can’t I can’t relate as much to an 8 year old as today I have a 10 year old and a 12 year old, but thinking of that 8 year old mind and trying to explain it to that age to me was really pivotal, and that is no longer a way that I’m really able to have us think as well about thinking, because that’s just so easy to go into a large language model and say, explain this to me as if I was 8. Explain this to me as if I was 15.

Bonni Stachowiak [00:10:07]:

I haven’t tried explain this to me as if I was 80, but that is another question that I would kind of ask. So that oh, no is an understandable oh, no, and faculty around the world have had similar oh, no’s that the approaches that we have attempted to Dave, to have there be integrity in the learning process, we’re bumping up against lots of things. And then you so that second thing I wanted to ask you about, besides your oh, no, is your comment about cognitive dissonance. What’s some of the dissonance that you’re seeing come up for people, specifically around integrity as teachers, integrity integrity as learners, and where is that cognitive dissonance showing up for you?

Eddie Watson [00:10:54]:

Yeah. Well, I think the initial oh no was indeed around academic integrity. Like, we are we are already having, especially as we went in and out of the pandemic and a lot of different systems being employed by students to get answers to things, you know, offer exams and and the like. And so we already had a new wave of academic integrity tools or challenges there that that some students were employing. And I felt like, oh, wow. So now we’re in this new domain. And then I think for many faculty, the arc that I’ve seen multiple times is that, okay, I recognize this thing exists. This is a problem within my class.

Eddie Watson [00:11:34]:

Oh, wait. We have plagiarism detectors. I think we can solve this problem. I can have this new technology be faulted by yet another technology, and then it’ll tell me. And then now I’ve solved this, and I don’t have to change anything around my my classroom practice. But as that’s unfolding, that’s that’s proving to be more more of a myth and a and a wish than the reality of how how these tools can be employed. So I think that there’s the challenge around, well, what’s going on in my classroom, and how can I ensure that my students are still achieving the learning outcomes that I that I aspire for them? So there’s just that first challenge around the classroom and how I can use different technologies and really just trying to ensure the integrity of my course. And a lot of first response has been, I need to keep this out of my class.

Eddie Watson [00:12:23]:

I need to make sure that my students aren’t using this. But there’s this companion challenge that’s created a sort of a new level of dissonance for higher education. And the companion challenge is that the world of work has rapidly adopted AI. So as we think about the challenges that we have with AI and trying to maybe suppress its use in our classroom, we also have a responsibility to prepare our students for the world that awaits them post graduation. And that world includes being able to use generative AI well, to have competency with it, to have AI literacies. So it’s almost like the classroom is kind of being ground in a way between 2 hard surfaces, with one being academic integrity and how do we ensure our students achieve the outcomes that we aspire for them. But then also, how do we indeed prepare them to be able to use generative AI and other forms of AI post graduation? So I think that what we’re seeing evolving rapidly is the notion of AI literacy as a as an essential learning outcome for higher education, regardless of discipline. So either within the major, within general education, or other programs or structures that you might have within your institution.

Eddie Watson [00:13:44]:

I I think figuring out how to prepare students for that world of work. I had a conversation just yesterday with someone who teaches mostly sophomores. And to think those students will not hit life beyond college for guessing, like, two and a half years at this point, Wow. Where will things be 2 and a half years? And how do you prepare them for that world that’s rapidly evolving? But I think it’s not going away from the world of work because there are efficiencies and cost savings that are being provided there. I was speaking at a conference in May of last year, and the panel before me was English faculty talking about their experiences in that 1st semester with generative AI. And I guess the story that really launched me into interest in this domain was and I guess this recognition of multiple levels of dissonance was someone saying that they had outlawed the use of AI within their class. It was a business writing class. And they had a student in that class who was a part time student, single mom, who had essentially a full time job and was coming to class part time.

Eddie Watson [00:14:58]:

And in her job where she did a bit of writing, her boss had instructed her, look, you need to save time with drafting of things. Just you can do a lot more if you’d start with chat GPT, produce the letter, and then you edit the draft. So you just partner with the AI, and you have a better draft, and you can do more work, and that enables you for more free time in other contexts of of that work life. So she was being told, you must use AI as a starting point in the real world. In the classroom, she was being told, you may never use AI. And she was very frustrated by that. She really wanted assistance in like, how do I how do I best communicate with AI? Can’t you teach me that as part of this business writing class that should be preparing me for the course or the work that’s waiting for me post graduation, or really not post graduation, but like when she goes to work tomorrow. So recognizing that dissonance and and that is sort of a false binary in that there are times that we want to see what students could do without technology or without AI’s assistance.

Eddie Watson [00:16:04]:

So a course should have both opportunities most likely depending on the larger outcomes for the course. But I think we do need to be leaning into preparing students to use AI within the discipline while also figuring out how to teach the learning outcomes and ensure ensure that students achieve the learning outcomes that we have within the class. I think that is the larger tension is that we do need to be leaning into this space, but how do we do it? How do we do it well?

Bonni Stachowiak [00:16:33]:

Tell us about what you learned even more about this when you talked to students about these issues.

Eddie Watson [00:16:39]:

Yeah. So as we were writing the book, one of the things that we as you might imagine for a book like this, we we did a lot of talking with faculty and had a lot of talking with students. And the student perspective, we’ve seen it sort of evolve where there was, you know, initially, I think students were like, wow. This is something that’s really powerful. And we agree. This should be outlawed from education. Like, we see that students who use this will have an unfair advantage against those of us that would prefer not to cheat. You are here to learn.

Eddie Watson [00:17:10]:

We we really do. I think the majority of students are those that, you know, are trying to learn within their courses. But students, their perception has shifted a bit, and it’s been largely around academic integrity. For instance, what we found late in the fall were students were not necessarily that concerned about being caught using AI or having AI detectors catch them cheating. Students we spoke to were largely concerned that they were going to be falsely accused of cheating based upon AI detection, which is largely a black box. I mean, we could talk about plagiarism detection, and there’s definitely issues and nuances around that. But at least a student often, plagiarism detection will bring back, this is the original article where this text resides. And here’s the student’s paper, and it highlights.

Eddie Watson [00:18:05]:

And so, you know, the faculty member then has something in hand that they could share with the student. And there can be a conversation around, so what choice did you make here? Did you just forget the citation? But there’s there’s evidence. There’s a smoking gun, you might refer to it. With AI detection, there really isn’t a smoking gun. It’s a black box where basically, students are being can be accused. And it’s very nuanced. Right? It’s not like this student cheated, this student did not. The AI detection tools typically report somewhat vague answers.

Eddie Watson [00:18:39]:

Like, there’s a 37% chance that this paper was written entirely by AI. You know, as a faculty member, I get that. What do I do with that? What’s my threshold of setting the bar? So students are concerned about being falsely accused, largely because they all know someone who has been falsely accused of using AI because of AI detection tools. And then again, it’s just a black box. So there’s just a lot of anxiety and fear. So I’m not sure that these tools help us in higher education. I think it might do more damage than good. There’s a lot of collateral damage.

Eddie Watson [00:19:17]:

Even the best tool on the market says that it gets it wrong one out of 20 times. You know, there’s a false positive. It’ll accuse a student of cheating who did not cheat with AI. And that’s the best in show tool. So I guess the question that I often post to faculty is like, what’s your level of comfort? Like, what’s your cutoff for student collateral damage? Is it like, you know, if 1 in 10 students get falsely accused, that’s okay. Is it 1 in 20? That’s okay. Is it 1 in 50? How about 1 in a 100? Is that would would that be if there was that tool doesn’t exist. But if there was a tool that got it right 99 times out of a 100, would would you accept that within your classroom setting? To my mind, I think the thinking about the student well-being crisis and also the larger concern, the diminishing faith in higher education, the diminishing confidence in higher education writ large, I think we might be doing more damage than good to students and to the larger industry of higher education by employing tools that falsely accuse a large number of students.

Eddie Watson [00:20:27]:

And then there’s questions about, well, who gets the false negatives? So not falsely accused, but who do we let through? It’s probably students who can afford to purchase the monthly subscriptions to the full versions of the AI and have access to the latest version of chatgpt4 turbo or the late the full version of Google’s Gemini. So do we do we really wanna use a tool that will really catch the Pell eligible students more often than the students who aren’t Pell eligible? There’s there’s so many ethical concerns within this space, and it’s just around AI detection. So there’s there’s there’s a lot for faculty to think. There’s a lot of cognitive cognitive dissonance just around the notion of academic integrity and AI detection. But then there’s also, well, what’s the alternative within our classroom? And then we have the what’s waiting for students beyond graduation that’s also placing pressure on us. So I I don’t think that I provided a really clear rubric of what I see as the variety of different tensions within, classroom settings right now, but there is there is an array for sure.

Bonni Stachowiak [00:21:35]:

Yeah. And I I as you were sharing, I’m thinking back to an the most recent time I had a chance to talk with Jesse Stommel, and something that he said continues to bounce around in my head, and that was, we have to know where our line is. And he was speaking ethically. We have to know where our line is. And I keep he keeps rattling around, like, various issues that I encounter. Where where is my line on this? And I think we have to be a little bit careful about where our lines are and and and and avoid that. I mean, these false binaries and what does it mean to cheat? What what does that mean? And and to what extent are we cheating students when we go with an approach that entirely limits any opportunity that they would have to use these kinds of tools in conversation with other people who are still in a formal learning season in their lives, you know, and and model for them what it might look like to ourselves be in a informal for some season in our lives, and and what does that look like? So but, I mean, it is scary because we do want to be able to have confidence that we are educating people well. It’s just to what extent is that policing people and to what extent are there other approaches we might be thinking about.

Bonni Stachowiak [00:22:52]:

I’m so grateful for these conversations. In the in the remaining minutes that we have before we get to the recommendations segment, I don’t I wanna shift to some more potentially cheerful topics. Are there cool things that you’re finding where you just go, oh my goodness gracious. Because you you talked about that, oh, no. Now tell me about, wow. What where are you just finding where you go, wow. That is absolutely remarkable of the kinds of things that you’re seeing it’s possible for it to do.

Eddie Watson [00:23:26]:

Yes. So I I I have those moments daily, and there are a lot of things that I’m really excited about. And I’ve always given my work, I’ve always tried to be sort of product neutral and not promote one product over another. But there’ve been a couple of things that I’ve just been really excited about. One is something called turbolearn.ai. And it’s a tool that students would have on their computers. They come to class and it listens in class and it takes summative notes. But then after class, it creates flip cards that students can then be quizzed on the material that the professor covered in class.

Eddie Watson [00:24:07]:

So it really provides an opportunity for students to kind of master the content that’s lectured on from one class to the next. Whoever built the tool, I don’t I don’t know the individuals, but it really has sort of an ed psych background, if you’re familiar with Henry Roediger’s work and the recall research out of Washington University in St. Louis. It’s like, well, this really aligns with that. I had I had one person I shared this with a couple of weeks back, and he was like, well, yeah, but aren’t my but isn’t note taking a learning opportunity? Like, aren’t we taking wouldn’t this take note taking away from students? And, yes, I guess we are replacing one task with an with a different task. But which one is likely to have the highest efficacy for student learning within that class? A lot of my conversations with students, they talk about class time being where they go to harvest stuff and that they learn the night before the test when they’re cramming and studying, and then they go and do the stuff and dump. They go to class the next day, and then they take the exam, and then they move on to the next thing. So is is note taking highly efficacious for learning? I think for some students, it can be.

Eddie Watson [00:25:15]:

For other students, it is collecting. But would it be more efficacious to then you experience the lecture, you’re paying attention, and then you have time to have have your have the lecture reorganized for you, re presented to you, and then quizzing you. And the good thing about such quizzing is that whenever you’re quizzed on, say, an hour’s worth of content, at the conclusion, you know which pieces you know and which pieces you don’t. So there’s efficiencies for the student. Okay. So I got these 14 questions right. I missed these 6. Well, now I know where to study.

Eddie Watson [00:25:52]:

These these are the things I don’t know. So it’s really kind of a mastery approach. So from a student learning perspective and something that is not an academic integrity challenge, you know, it’s just a tool to help students capture content and then to revisit and better understand and learn the content. I like that tool because of those characteristics. There’s also this tool called AI Tutor Pro that just within the last hour, I had someone provide me with a demo. And in addition to that tool I’m Bonni pull it up on my screen so I can make sure I’m describing this well. There’s this group in Canada that they’re funded by the government. So this tool is entirely available to everyone, entirely free at this point.

Eddie Watson [00:26:37]:

There’s no business model. The business model is we’re sharing this with higher ed. So I like that about the tool number 1. Again, sort of my open background, really very much liking that. But AI Tutor Pro, and they also have something called aiteachingassistantpro.ca. And these tools, one is a quizzing tool. Like, you tell it which domain of knowledge you’re studying, and it will Dave you it’ll just quiz you about it. It’ll ask you questions, and it will help, help you sort of Dave you feedback.

Eddie Watson [00:27:09]:

So it sort of helps you better understand any domain. So, you know, if you’re going to be visiting Paris, you could have it quiz you about Paris. So that’s that’s one thing, again, from just sort of like a learning a domain area. And they’re they have partnered with OpenStax. If you’re familiar with the OER world, they have built an LLM with all of the OpenStax content. And so when you use the the the tutor tool, it will give you the citation and the answers that it provides, and you can click on it. And it’ll take you right to OpenStax and show you which page within that book that content exists within. Attribution is one of the things that’s a real struggle or challenge with generative AI, certainly.

Eddie Watson [00:27:52]:

But they have a solution for that within the the open world. So it’s a great tool that I think they’re Bonni to have that to market in the next month. Think that was the timeline, but he gave me a demo. But the other one is AI Teaching Assistant Pro, and it will help you create test questions. It will also you can give it sort of like a domain. It’ll write an essay prompt for you. And if you’d like, it’ll provide you with a scoring rubric with the attributes of what excellent work looks like versus other elements of work. It also will help you create a syllabus for a course.

Eddie Watson [00:28:29]:

So these are tools that are designed to help faculty do things more quickly and and really, in many cases, better. Just as as an aside, except for the one about OER, the brain behind these other tools is indeed ChatGPT 4 turbo currently at the moment. So so you that’s that’s, where they’re pulling data from. But that’s really exciting to be able to help us build Syllabi, help us write test questions. Ultimately, it’s not replacing us. I mean, we’re the last arbiter of quality. We have a co author now, but we’re first author. We we make the decisions regarding what we bring to our classrooms, what we share with our students, but it’s a tool to help us.

Eddie Watson [00:29:10]:

It’s a brainstorming partner. So this kind of notion is what really excites me about generative AI within the teaching and learning context these days.

Bonni Stachowiak [00:29:19]:

As I I mentioned about curating slash writing this open textbook for the course, and I have used chat gbt 4 to write some questions. And it’s one of those things where when I look back, this has this has been a very difficult endeavor. I had not planned on ending up with over 70,000 words in this open textbook, and so the questions are not perfect. I could certainly go back and refine and improve them. I’m oftentimes going through them with the students. That’s we’re using it as a means for review as we’re doing the class together. And, you know, sometimes I find they’re too easy. The the right answer is very obvious to discern.

Bonni Stachowiak [00:30:00]:

But I also think about I I haven’t used a for profit textbook company in products in a long time, but I’m positive they were using similar things to develop their questions. They also were not exemplar as far as good questions. So I think we have to be careful thinking about what we’re comparing it to. And I know that there would be all these ways in Excel you used to be able to Dave, you know, here’s the language, and then we’re gonna we’re gonna use different formulas to replace the variables so that you can vary the types of questions. You might, you know, be able to expand a test bank to be twice as large by using tools like that. Well, that is very similar to what’s happening in these large language models when they’re doing the same thing for us, so I think that’s important. And I wanted to go back to you had said earlier in our conversation. Do we really want to be discriminating against who can afford the best tool, the paid tool versus what’s available for free.

Bonni Stachowiak [00:31:00]:

And I felt like when you were talking about some of these note taking tools, I’m thinking back to the research I’ve seen from, for example, Michelle Miller talking about the ways in which some of that note taking literature is we should really be questioning that. Specifically, I remember her talking about handwritten versus typed notes that it’s far more nuanced than what the headlines at the time would tell us. And secondarily, we know that one of the main accommodations for students with learning disabilities is having to do with note taking. And I can tell you from having seen what it’s like to ask another student to take notes in a class or ask it is truly remarkable. Is it And and, I mean, it is truly remarkable. Is it perfect? No. But neither were the other means of having access to notetakers. So this is kind of I don’t wanna overly idealize this, but it is it it contains some avenues for leveling the playing field when there’s a lot of avenues that are most definitely not in this space.

Eddie Watson [00:32:15]:

Yeah. And it’s it’s not an either or. I think I kinda presented it that way. It’s not like the student can either take notes or the computer can take notes, but the computer could take notes and the student could do what the student has always done in a classroom, which is listen, take notes as best that they can. And now they’ve got a companion that will be quizzing them. And to be honest, you you if you realize, oh, I missed these questions and it’s not in my notes, it might highlight that I’m not the most efficient notetaker. Conversely, I might feel confident. Like, oh, wow.

Eddie Watson [00:32:42]:

I do a really good job of taking notes. I feel better in other contexts where maybe this tool isn’t appropriate. But, I mean, it’s there are there are definitely nuanced new ways to leverage these tools that will indeed help learning and that aren’t a threat to academic integrity.

Bonni Stachowiak [00:33:00]:

My husband, Dave, recently had asked Chatt GBT to analyze some responses that he had gotten about challenges that leaders face, and that got me thinking. And and so I was going back to a conference I had been able to attend, and they had they had taken the poll responses and posted them on a publicly available website. And they had anonymized the data, and so I thought, I wonder what it’ll do to say, you know, could you analyze one of the questions I remember the poll that they had taken had asked, what sorts of fears do you have about what AI is gonna do with your jobs? This was for people who do instructional design in their roles, and it was truly remarkable to see how if I had just gone through and looked at these poll responses, which I do many polls when I when I engage with faculty from all over the world. And and then to to see how well it synthesized that, I mean, it it’s you said, oh oh, wow. I mean, I I’m I’m just joining you in in thinking through these possibilities. And as you said, I am every day too just, I am still every day going, oh, no. And we’re about to transition to the recommendation segment where we will do some oh, no together, but I’m also a wow. I mean, truly, truly some incredible features that we can only hope can truly make education more accessible for all.

Eddie Watson [00:34:25]:

I agree.

Bonni Stachowiak [00:34:26]:

Alright. This is the recommendations segment, and I would like to start with an oh, no. This is an article entitled dialect prejudice predicts AI decisions about people’s character, employability, and criminality. And so as we look at the extensive use of people, we we talked about in the workplace, you know, using these large language models, A significant drawback in this study is the problematic ways when we analyze groups, for example, African Americans and some of their dialects. This this set of researchers found that there are, then more likely to show up as having differences in terms of employability, judgments, and assessments about people’s character and criminality. So we wanna be definitely continuing to learn about this and to find ways to influence that we can mitigate those concerns, and this is an ongoing concern, of course. And there are a couple of things I wanted to share where I’m really enjoying the conversations. I’m I’m part of a listserv called AI in Education.

Bonni Stachowiak [00:35:41]:

And so Leon Fertz wrote the myth of the AI first draft, and he spoke because I I had been asking this with prior guests about what’s the benefit of having a blank Dave. And we know that as writers, most of us that have sat down with a blank page have had those struggles, and to what extent are those struggles beneficial for us? It’s beneficial to be bored. It’s beneficial to struggle. It’s it’s creative creativity can emerge from those spaces, so I appreciated what Leon had said about the myth of the AI first draft, and he’s already agreed to come on the show, although it’ll be, as of you listening to this, it’ll be a few months because our schedules were a bit off as far as time away, but I’m looking forward to that conversation with him. And then Jon Ippolito also engaged in that conversation and wrote an article called how to teach AI and still put people first. And he had what he called an AI sandwich, which cracked me up. And so the idea he broke it up into 3 phases where stage 1 could be AI as brainstorming partner. And, Eddie, of course, you and Jose write about that and give some prescriptions in your book on this as well.

Bonni Stachowiak [00:37:02]:

How can we partner with AI in this way? And then for him, stage 2, human to human knowledge creation. And the example that he gives in the article is that the student might go, and in this particular research project, might go visit their hometown and interview her mother, her grandfather, local business owners, members of the town council, anyone with a stake in the research question. And this reminded me a little bit of a colleague at my university, Mike Jimenez, and he always does such an amazing thing with the seniors. He’s a history professor, and the seniors will come in, and they will have interviewed a family member, but it’s also helping them use a historical lens in their analysis. So I was just really thought that blend was very intriguing to me. And then finally, step 3, AI as a writing paper, which, again, I wanna emphasize that that Eddie and Jose talk about in the book as well. And, Eddie, what do you have to recommend for us today?

Eddie Watson [00:38:04]:

Well, I think it’s really kind of a a path into AI. I guess one of the misconceptions that I sometimes hear from faculty that are just entering this space is the notion that AI is kind of monolithic. Well, it’s ChatGPT. That’s the thing that we hear about all of the time. So my recommendation is typically, like, if you if you’re if you’ve been working with that a little bit, try out some of the other AI brains that approach the challenges that we might provide them in different ways. So there’s there’s different worlds. There’s there’s Gemini, there’s ChatGPT, there’s Claude, there’s probably 6 or so out there that’s within this space. So give them all a try and try the same prompt in different AI, and you’ll see that you’ll get very different answers.

Eddie Watson [00:38:52]:

Sometimes you get very different answers. Recently, I was I had to do sort of a workshoppy kind of description thing, and 2 of them were very close and the third I typically try 3 whenever I’ve got something that I want. I want to be able to respond to and sort of, like, synthesize, like, okay, 3 different ideas, and then I’ll make the final decision or I’ll synthesize things. But one was very different. Often, Claude, I find to be sort of the better brainstorming partner, though it hallucinates maybe more than, some of the others. So play with more than just a couple, and also just bite the bullet and pay for a subscription to a couple of these tools to better understand what your students, or really a subset of your students, are indeed using in the classroom. The paid version of ChatGPT is far better than ChatGPT 3.5, the free version. There’s there’s notable differences in its performance, just how fast it is, versus the quality of response that you get as well.

Eddie Watson [00:39:54]:

So that that’s one key recommendation as you sort of enter this space and begin to better understand this world. Also, there are a number of really strong classroom based faculty who are becoming thought leaders around how to be creative with AI. So there’s a number of, well, they certainly have presences within LinkedIn. And so this is a great place to start is with LinkedIn because they’re typically posting from their their own blog or their own substack or whatever it might be. But Lance Eaton at at College Unbound is someone who does really interesting work there. I think you mentioned Michelle Miller earlier, who’s at Northern Arizona University. She does some really great things, and you can subscribe to hers and get get a message in your inbox. Anna Mills is another person who’s been doing really great work.

Eddie Watson [00:40:41]:

She teaches writing. And so for those of you that are really focusing on the challenges that you might be experiencing around writing and sort of creative ideas to do things differently around writing. She’s someone else that’s that’s great to go to. But that’s, I think, a great sort of path in is here here are other people that are are are really posting, couple of times a week, interesting approaches and what they’re finding. That’ll kind of help you stay sort of plugged in. The landscape around these tools is definitely changing almost daily. Apple hasn’t even come to market yet, but we know that they swallowed up, like, a dozen different AI companies in the last few years. So we’re anticipating something reasonably spectacular regarding, features and performance this summer from Apple.

Eddie Watson [00:41:31]:

So this just it’s just it’s just continuing traffic. There’s about 2 years, and then we’ll start to see that some of these companies will be swallowed up or some ideas will merge. But it’s, definitely I think the the ride that we’re on currently will extend at least another year or 2.

Bonni Stachowiak [00:41:48]:

Eddie Watson, it has been so great to reconnect with you today, and I so appreciate, especially, that you can give us the start to people that we might be able to continue to have these conversations with even if it’s only in our own minds. I feel like you and I have had many conversations in recent years, but not actually, you know, a two way thing. It’s so fun to have gotten to talk with you today, and thank you for this incredible book. I was so honored to be asked to read it in advance and give some feedback, and I thought that you and Jose absolutely hit it wonderfully well to be bringing up these real ethical issues while at the same time balancing out those ethical issues with what is it that we need to rethink about how to teach, how to learn, and how to work? And I I feel like we just got started, so I guess I guess this is gonna have to be it it can’t go this many years before we talk again because there’s so much that we can talk about here. Just thank you so much for your time today and for this wonderful book.

Eddie Watson [00:42:45]:

Well, thank you for your kind words, and thank you for the invitation to be back on. And I I look forward to the next opportunity to chat with you.

Bonni Stachowiak [00:42:54]:

I’m not sure if it’s gonna show up in the recording, but I’ve been down with a little bit of a cold. And this conversation with Eddie Watson really brought me all the life and energy, and I’m so grateful for him coming back on teaching in higher ed. Today’s episode was produced by me, Bonni Stachowiak. It was edited by the ever talented Andrew Kroeger. Podcast production support was provided by the amazing Sierra Priest. Thanks so much for listening. And if you’ve yet to sign up for the weekly teaching in higher ed update, you’ll receive all of the links. And there are going to be some really good ones in in this one that come out in the most recent episodes show notes as well as some other resources that are new and distinct for the emails.

Bonni Stachowiak [00:43:39]:

So head on over to teachinginhighered.com/subscribe, and I’ll see you next time on Teaching in Higher Ed.

Teaching in Higher Ed transcripts are created using a combination of an automated transcription service and human beings. This text likely will not represent the precise, word-for-word conversation that was had. The accuracy of the transcripts will vary. The authoritative record of the Teaching in Higher Ed podcasts is contained in the audio file.