EP 55: How to properly leverage AI in the classroom
11 Jul 2023 · 25 min
Jordan Wilson and Kelsey Barringer, CEO of Packback, discuss why banning AI from classrooms is misguided. Instead, instructors should set clear expectations, model appropriate use, and redesign assessments toward authentic work. AI detectors are unreliable and shouldn't be the sole evidence for academic misconduct.
Questions this episode answers
July 11, 2023.
An AI-powered writing and discussion platform for higher education (expanded to K–12).
On servers in the United States and in Canada.
About 15% of the grade (discussion use case).
About 18 months prior to the episode.
Chapters
This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.
Should AI not be used at all in education or should it be used everywhere? Those are some of the things that we're going to tackle today on Everyday AI. This is a daily livestream, podcast and free newsletter helping everyday people keep up with AI. There's a lot going on. Let's, let's be honest. You could take a week off from reading AI news and everything has changed. So that's what everyday AI is here for, and that's what we're going to do today. So before we have a great guest come on and talk about AI in education, let's first talk about what's going on in the world of AI news. So, so now we're detecting AI deepfakes with AI. So a new report looks at how the Pentagon is using an AI deepfake detector. So kind of, kind of crazy that the deepfakes are obviously being made with AI, but we're using AI to detect them. I'm not sure how that works, so make sure to check that out in the newsletter today. Another one. This was somewhat weird if you haven't taken a look at it yet. So there was a press conference given by robots yesterday at the AI For Good conference in Geneva. So pretty interesting stuff. But don't worry, they promised not to kill humans and they said they wouldn't take our jobs. But it's pretty funny. We'll link to it in the newsletter. They did give some strange looks when we asked them some of those questions. Last but not least, our third big AI news story for the day. So Nvidia, their AI market share is soaring. So Citi said that Nvidia will have 90% market share in AI chips. That's wild. So we've talked about this on the show before, but here's why chips are important specifically for the economy. Because all of these generative AI tools, they all require these GPU chips. Nvidia and a couple other companies here in the US US is really the leader in making these chips and sending them, you know, exporting them overseas as well. So there's been a lot of discussion on should the us, you know, be sending these chips to other countries that might be using them to either bolster their economy or to use them in war? You know, we talked about that yesterday, kind of previewing a Netflix documentary that came out about that specific thing. So with that as a reminder we are live. Rastafa, thank you for the comment. Already saying good morning. So if you are checking this out on the podcast later in the day on Spotify, Apple or wherever you get your podcasts, make sure to join us. We do this live every single weekday at 7:30am Central Standard Time where we bring on amazing guests. So speaking of amazing guests, let's bring one on right now. So we have with us today Kelsey Barringer, the CEO of Packback. Kelsey, thanks for joining us.
Hey, thanks for having me.
All right, so quickly, just tell everyone what Packback is because there's a lot of. And again, Packback. I almost said it wrong. You know, I almost say backpack. But tell us a little bit about Packback and what you all do in the education space.
Yes, fortunately or unfortunately, we are often autocorrected to payback Pacsack many different things. But what we are high level is an AI powered writing and discussion platform for higher education mainly. But we also just entered the K12 space about 18 months ago. The way that you can think of us at a high level is for students. What we do is offer AI assisted feedback, writing assistant and also research assistant. We give them feedback on the research that they're doing. And then for instructors, we're offering some grading support. We don't want to do 100% of the grading for an instructor because they're ultimately the content experts, but we do want to support them with grading things like grammar mechanics so that they don't have to be a red pen and they can truly be content experts.
Yeah, that's great. So, so more or less it is. It's AI assisted tools for both students and for educators.
Yep, absolutely.
So, so I know that you're a former educator yourself. So what are your, you know, if you take off the, the CEO hat for a minute, what's your take just on AI in education in general, Is this something that, you know, maybe if you would have told yourself five years ago that you would feel the same way you do now? Or what's your, your take if you take out, like take out the, the role that you are right now?
Yeah, absolutely. So I think more more broadly, when I first started teaching, I was teaching really large core sizes. I was teaching high school chemistry and I had 36 seats and 46 students in my classroom. It was, it was bananas. And I was 23 myself. So what I did was I used and I embraced technology to solve the problems that I felt like I needed more support on. So I used Kahoot. I was a huge fan Of Kahoot, I built my own website. So in general, I think technology as a whole is amazing at solving problems when used correctly. Right. When it's used poorly, it adds stress, it adds inefficiencies. And I think of AI the exact same way. AI is just really wonderful, magical technology that can be leveraged as a tool. And when used properly, it's extraordinary. And it saves instructors time. And then they get to spend more time doing what they want to do with their students, which by the way, is not like grading them on spelling their incorrectly in this context. So I think AI in education has its place and it also has best practices and worst practices.
Yeah, yeah, for sure. Speaking. Let's just jump straight there. So speaking of worst practices. Right. So I know that students are not always using AI as a compliment or a supplement. They're using it as a replacement. But then on the same side, I think you have professors maybe using AI incorrectly. You know, kind of the famous or infamous story is the UT Austin professor that decided to fail most of his students because he declared that they were using AI when you can't really declare that. So what's, what's just your thought on kind of that, that, that line of where AI should be used versus okay, it's, it's time for, for humans to, to get involved here.
Yeah. Yeah. Well, let me just start this by saying my heart really does go out to the professor and the teacher right now, especially the professors who did this job because they love teaching their discipline and want to be connected to the students. So I want to, I want to preface with, I have the most sincere empathy for instructors right now, but I think one of the most misguided reactions we could have to generative AI is banning. So we, we can't just see a new technology and say, it doesn't have a place in my classroom. Maybe it doesn't have a place in your classroom. Like, maybe it would be really inappropriate to use AI. But I think it's so important for instructors to explain to students why it would be inappropriate. And if the reason is because the instructors just don't want students to cheat, we have to dig deeper. We have to turn inward and self reflect. Is this really about students not cheating? Is it about control? What is this really about? But I think banning is misguided because students are using it, they want to use it, and they want to hear from instructors about what's appropriate, what's not appropriate. Because 80% of students don't want to cheat. Right? They don't want to Feel icky. They want something out of their education. And then just again, maybe putting my CEO hat on, but also my former teacher hat. I do not think technology will ever fully replace some of these different careers like teaching because so much of teaching is being able to like, connect with a student and differentiate content for that specific student, have a face to face or zoom to zoom human interaction and inspire that student, challenge that student. Something that I just think humanity needs to be there for. So I don't think AI should be holistically replacing an instructor workflow. I think it should be supplementing pieces of it because most teachers are not working less than 60 hours a week and it's unsustainable.
Yeah. Yeah, for sure. All right, we already, we already have some great questions, so I'm going to get to them one or two at a time. So, Nancy, shout out Ozynga crew. So, Nancy saying, can you bust the myth that teachers know if students use AI for writing assignments? Kelsey, what's, what's your take on that one?
Yeah. So AI detectors are not very accurate. So my advice is if you are using an AI detector, you have to do research before you make any claims against that student. So the, the best case scenario, and if not every educator could do this, my recommendation would be first day of class, have them do a written writing assignment that would establish their unique voice and then use that as your data point to juxtapose a writing assignment that maybe got flagged for AI detection. Or you're kind of like your alarm bells are going off at the end of the day. I think educators who've been doing this for quite some time have been grading writing for quite some times, understand the holistic student voice and what the average student is going to say in a piece of writing. And when they get like this really wonderfully written, expert written piece of paper from a student that maybe doesn't show up at class, your alarm bells go off again. Do research, take it with a grain of salt, compare it to previous writing samples, compare it to previous writing samples from other students, and then have a conversation with the student. But I think my, my main takeaway is you can't take detection at face value. It's great. Going to be a dicey place to live. Yeah.
And we did, we did chat pre show about this briefly. Yeah, My take is Nancy, no, they're all it. Like, I haven't used all the enterprise solutions, but anything commercially available I've tried. Very easy to bust. I won't say they're, they're nonsense but yeah, they don't. They're meritless, I'd say. I'll say that like, like, like that. So, follow up question, which is great and I think this is could talk about this for hours, but another question, is using AI for class assignments always considered cheating? Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everyday AI.com partner to get in contact with our team or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on GEni.
The answer is no. I've been using ChatGPT to help me create outlines and agendas for this webinar series that I'm running. And I'm the webinar is my own words. I'm, I'm putting together my own slides, but I'm using AI to help me just with the outline and to organize my thoughts. And I think a lot of us would agree that if a student was had to write a 5 to 10 page paper and ask GPT for feedback on their paper, ask GPT to help them with an outline but then wrote their own words. I don't think any of us would consider that cheating. I think we'd say that they've found a really efficient way to not have to go to the writing lab and maybe just use GPT at 3am Right. But ultimately I think this is where my recommendation to so many instructors is set your expectations at the start of the term and reiterate your expectations. And if a student blatantly disrespects an expectation you set, then yes, they're cheating. But you know, I think it's, it's our duty as educators to guide students and how to use these tools because if we don't they're, they are going to misuse them at the end of the day.
That's a great point. And Kelsey, even going back to the original, setting the expectation before and using that as a comparison I think is a great tip for educators and just for students to know too. Your teachers, professors, TAs, they know that you're probably using it. So I love that concept. So another, another question here. So Rastafa is asking about Packback. I think so can you share tools across users? Do professors on the content? So yeah, tell, tell a little bit about the actual platform and how that works in terms of ownership and access.
Yeah, well the platform itself is mainly used in higher education. Instructors adopt Packback as a tool in their course. Kind of like they would adopt a technology like Clickers or Top Hat. And we're usually a portion of the grade because we are an online discussion platform. See us about 15% of the grade and then also a writing platform for long form. That could be upwards of 100% of the grade. Right. It's all student driven inquiry based discussion or student written work. So all of the information that's in Packback, all of that data is owned by the student. That is their voice, that is their work. And we house it in Packback. We have servers in the US and a separate server in Canada and right now we're only US and Canada, but it's their work. The instructors are obviously guiding them on how to use it and the instructors are grading the content, things like that. But it's student owned work.
Yeah, great, great, great thoughts. And we have a couple, couple more comments I get wanted to get to. So yeah, Dr. Castro, thanks for joining us. Saying I think we need to redefine how we teach and how we test. Absolutely. You know I, I, I've said this on the show many times. If I was a professor there would be a 10 to 20 page paper due every single day in class. Brown win. Good afternoon from South Africa. Love South Africa. I'd love to go back. FYI. Another question though from Audrey. So, so Kelsey, where do you draw the line? Student input versus GPT powered work? That's a great question and I have no clue how you can provide a good answer. But that's, that's what you're here for. That's why we bring on the experts. So yeah, where does that, where do you draw the line?
Yeah, it, that's such a good question. It's like I'm already seeing in a few years like Supreme Court cases, right. Of like students challenging you Know, I'm just seeing it all happen. And, and it's not a line. It's like this very thick, foggy area. And I don't know if I have the best answer for this. I think I just keep going back to if I were an educator right now, which I used to assign a lot of writing, specifically lab reports. So it wasn't like creative writing, it wasn't argumentative writing. But what I would be doing is every single assignment reminding my students this is the appropriate way to use GPT on this. This is the inappropriate way, and I would probably model it for them. And then if I felt like they were violating these guidelines that I set up, I would have a conversation with them. And it's pretty easy to tell like that. When I've talked to instructors about how they're approaching these conversations, it's like, all right, tell me about how you came up with the idea for this paper. All right, how did you come up with the outline? What about this? I thought this was an interesting point. And if a student can confidently say like, this is what I thought about this, they can explain their process. They probably wrote that paper themselves or at least did it with very little input from GPT. They probably used it properly in the same way that adults are using it to bolster their careers. Now, if a student looks at you wide eyed, you're probably like, all right, let's, let's talk about this. Why did you use GPT? You know, and then you have a conversation with them. And again, most students are not trying to cheat because they're, you know, ding dongs. They're, they're and bad kids. They're usually cheating because they're working multiple jobs and they're busy or they're in this assignment because they are in this course because they have to be here. So help them understand why they did it and try to address the behavior. But I think it really goes down to what's the, what's the true relationship and purpose, the purpose of the relationship between the student and teacher. And let's like let that inform the line and not. I don't know if that answer works, but that's kind of how I feel about it.
Yeah, I think with some of these things is there's no right or wrong answer yet. Because I think for a while higher ed especially is going to be operating in this gray area. Because, you know, whether students admit it or not, I'd say the majority of students are using GPT to write their papers. I've heard from them they're, you know, you even see these photos, you know, people who graduate and put the ChatGPT logo on the top of their graduation cap. Right.
That's amazing.
Yeah, yeah. But as, as a former educator. Yep. And I'm sure you still have, have friends who are still in the field. How do current teachers view CHAT GPT? What are their thoughts? Are they totally against it? Is it like, is it like the dark side, you know, Star Wars? What are, what are their thoughts? It's obviously maybe different for you now because you're pushing the technology and advocating for it, but what are teachers now viewing the ChatGPT technology as?
Yeah, it's a huge spectrum. So I've been leading these series of workshops and webinars with the League for Innovation and the Community Colleges. It's a wonderful association that basically advocates for transformative approaches to pedagogy in community colleges and will survey a thousand attendees of a webinar. And it could not be more varied. It's like 25% want to ban it. They're super scared. Student voice is dead. Long live education and critical thinking. 25% are like, we're requiring it in all of our courses and all of our assignments. My students have to use it in order to pass. And that middle where is like the, the, the majority of, I think our voices are. I know it's important that my students learn how to use this correctly. I know it's important that I learn how to use this correctly. As an educator, I don't know where to start. I don't know how to move forward. So I think that's where the majority lies. But it's still a very broad spectrum. And you know, education tends to move slower when it comes to adopting technology. Like my, you know, my first classroom, I was bringing in all the technology personally. And my biggest fear is that we'll, we'll just kind of sit back and wait or like institutions will be like, well, we need to, we need to do a limited test study, then we need to run research. And then five years later they're ready to embrace AI in classrooms, which is like, we're just going to have to be a little bit more, I think, like, listen to our guts instead of take the time to inform, like let data inform our decisions over years of research. Because that's tends. Educators tend to be like that. They want to use, you know, research back strategies in their classroom for sure.
Do you see with, you know, and obviously so many classes are, you know, online classes now, so people are on the Internet. On a computer. But do you see in the near future a rise in kind of old school education? More handwritten assignments, more tests that may be, you know, fewer papers, more tests. Do you think education is going to be shifting in any of those ways?
Oh, I think a lot of instructors have bought, brought back blue books and will. I would caution against it. If I was asked to write a five page paper right now by hand, I would actually feel like physically sick after writing because I haven't done it in so long. I journal, but I'm not journaling a five page essay. It's been so long since we've had to do that. So just, I'd be mindful of that now. I do think this viral rise of ChatGPT is forcing instructors to really reflect on the writing assignments they're doing. And if it is easy for GPT to get an A on your paper, you need to rethink your assignments and your prompt. So I just think that it's going to challenge what we've been doing and force us to build more authentic assessments. And like, for example, hey, I want you to go out into the world, do some research. I'm. Instead of you doing a 10 page paper over the course of eight weeks, I want you to actually come into office hours. You're going to give me five minutes, you're going to tell me about the progress on your project and at the end you're going to actually present to me and then the paper will just be part of it. So it's, I think, more like holistic assignments that really force students to act and learn and apply and present and engage. I think that's what's going to happen. I think it's also going to force us to be much more dynamic as educators and change our assignments more often and update them. That's also a good thing. But it does put a lot of work on educators. So that's just, I think something to be mindful of is constantly reinventing the wheel is a lot of work.
Yeah, yeah, absolutely. Yeah. Not, you know, personally, I do think I feel for the teachers. You know, I have a, I have a cousin who was a teacher for years and just got out of education and I can, I can see why. You know, the, the jobs I think are getting harder, the pay is not necessarily reflective of the increased demands. So yeah, it's, it's definitely, it's definitely a space and a sector that is facing challenges and the AI, you know, piece doesn't help. I don't, I don't think so we've talked about a lot. We've talked a little bit about Packback and what you all are doing. We've talked about, you know, some of the challenges and some of the, you know, features and benefits that this, that this new technology brings. But let's, let's end with this. So another great question. So Nancy saying criticism she here is that students will lose critical thinking skills. So let's end with that one. Kelsey, what's, what's your, what's your take on that?
Yeah, so my take is some of the smartest people I know are using GPT often. And so I'm like, I, I think I push back on this concept of technology forces us in the direction of less thinking and like less application and actually probably helps us get rid of tedium and noise and actually level up what we're doing now with students. Yeah, there is a, there is a total possibility that when a student goes into a gen ed course they use GPT for all assignments and then they don't do any application. They don't critically think, they just GPT every assignment. So that's possible. I think it's going to again just challenge the way in which educators teach courses that maybe are required at the university level, are required for the major, are challenging or hard or intimidating. I hated reading anything that wasn't written in this century because I found it really challenging and if I had GPT to help me digest the content, I would probably be less scared all the time. So yeah, I think we are at risk of students not authentically engaging in courses sometimes.
Yeah, absolutely. I'll, I'll throw in as we, as we close out, I'll throw in my take. I think it's hard. I think the more and more we use and we see the benefits of this AI technology, I think it is harder, you know, having more and more easy access to a world of information even faster and more, you know, in bite sized pieces. I do think it hard like critical thinking, Nancy, I think is going to be hard, but we could talk all day. We won't though. Thank you Kelsey, coming on sharing about Packback. Thank you so much for taking all of our great questions and really appreciate you coming on the show.
Yeah, thank you so much for having me.
All right, as a reminder, if you liked some of the things that you heard from Kelsey, if you want to check out Packback, if you want to see some of the other resources we are talking about, make sure to go to your everyday AI dot com. That's your everyday AI dot com. Every day we sent out a newsletter breaking down the conversation that we had live on the show, sending more resources, so make sure you check that out. Also, we did just launch our new giveaway, free six months of ChatGPT plus and private training, so make sure to check that out. So, Kelsey, thank you again for joining us. And we hope to see all of you back tomorrow and every day on Everyday AI. Thank you.
And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit your everydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.
More from LARRUZ


Ep 881: Faster AI Agents, Fewer Human Coworkers: The Overly Productive Future of Managing Agents? (Replay)


Ep 879: New: OpenAI Presence. Has The AI Customer Service Takeover Finally Arrived? (Replay)


Ep 878: What’s Coming Next: 5 AI Trends, Problems and Opportunities around the Corner (Replay)




