Watch

Listen

Read

 ChatGPT, can you write me a thousand word essay on the effects of ChatGPT in the classroom?

Is that you just saw? Someone trying to cheat their way into a good grade by using ChatGPT by turning in an assignment that they didn't really do? Just a year ago, that was unheard of, but in 2023, ChatGPT and AI is all the rage. I'm going to take a dive into why there's so many students using ChatGPT.

The integration of ChatGPT in classrooms has sparked both excitement and concern among educators and experts. While it offers undeniable benefits, such as instant access to information and interactive learning experiences, concerns have emerged regarding its potential drawbacks. One major concern revolves around the possibility of over reliance on chat GPT, which could undermine students critical thinking skills and hinder their ability to seek guidance from human instructors.

Additionally, the risk of misinformation and biased responses generated by the model raises questions about its accuracy and appropriateness in an educational setting. Have you seen a student or know someone that has tried to turn in work that was generated by ChatGPT? According to SocialScienceSpace.

com, among the 216 students who use the tools for school, half of them use AI for some parts of their work. But do the majority on their own. 30% use AI to complete most of their assignments, but make revisions. 17% turn in work completed by AI without making edits. And 3% preferred not to respond. Joining us to talk about this is Caroline Scott, who teaches college composition.

This is for a personal narrative assignment. So they were essentially using it to make up hypothetical stories about themselves. And it was just, it was really obvious in that case that it had come from GPT. And I, I wasn't confrontational about it. I just said, Hey, this is pretty obvious. I think the way I phrased it, this was tagged for AI content.

I didn't say it was tagged by my brain, it it came up as AI generated and Another teacher joining us is Garrick Wilhelm, who has been in the tech industry for 15 years and has taught for two years. There's definitely been work that we suspected, but then it's nearly impossible for you to go and check by putting that information into anything and coming out with the same result.

output because even if you put the same terms in at one time, you're going to get something out at a different time, different out at a different time. So it's difficult to have a definitive check, but we know when work isn't the student's voice, when it's outside their normal vocabulary, when it's it's outside the knowledge spectrum of what we're learning.

Also joining us is Joe Nesbitt. He has been an educator for 25 years and is a instructional technologist. End of the school year, and literally the last week of the school year the topic first, and it was coming from the English department, and so I'm certain that there, there have been some students who have utilized the tool, we'll put it that way but I have not seen it firsthand just yet.

Is AI going to be a problem? This has been subject to debate since the launch of ChatGPT. The main concerns surrounding ChatGPT and AI center around three key areas. Firstly, there is the worry about the potential for biases and inaccuracies in the generated responses, as AI models like ChatGPT are trained on large datasets that may contain inherent biases.

Or misinformation. Secondly, there is the risk of over-reliance where students may become too dependent on AI for answers, potentially hindering their critical thinking and problem solving skills. Finally, there are ethical concerns such as privacy issues related to data collection and the responsible use of AI technology.

Is it a problem? Yes. Is it a problem? We can't. No, not at all. It's something we can easily work with, and I think it has many valuable uses also in the classroom. Short term, it's going to be a significant problem. I think in the long term, hopefully, more understood, it will be a tool that will be embraced.

Because it's not gonna go away. I would imagine the same or similar type of fear, panic conversation was occurring when calculators were being put into the classroom, right? It's the same sort of thing. So it's not gonna go away and we have to figure out how can we use this?

How can we police it? But how can we use it? Because eventually there's going to be careers that, people will be able to go into because of AI and things like chat GPT. If you keep teaching the way you've always been teaching and you don't change your curriculum and you don't change your approach, your teaching method is going to be completely ineffective.

Like you're going to have to change everything as a teacher. And I get why people don't want to do that. I don't see it as a problem. I really don't.

A teacher quoted anonymously. And the study. com survey shares, I love that students would have another resource to help answer questions. Do I worry some kids would abuse it? Yes. But they use Google and get answers without an explanation. It's my understanding that ChatGBT explains answers. That would be more beneficial.

Forbes also compares ChatGBT to the calculator. Saying, when calculators came into the classroom, many declared that the decline of math skills would soon follow. Research conducted as recently as 2012 has proven this to be false. Calculators had no positive or negative effects on basic math skills.

According to EducationWeek. com, Many teachers more likely to permit than prohibit students' use of chat. G P T I think I will be a fan of it. I'm pretty new to it as well. Just poking around, playing around with it a little bit. It's fascinating to me. I don't, again I think back to when I first got into this position YouTube was the scary thing.

Should we block YouTube on the school's network? Should we allow students to use YouTube? And now granted that there's some other things there, with. inappropriate videos and things like that. But but it's similar in that, hey, now we couldn't have gotten through the pandemic without you for education, and tools like that. I'm an. And that I think that there's a place for every tool in education but it all comes down to efficacy and how it is utilized. Make a presentation with gamma. And they really enjoyed that because it was done in two minutes and you use their, I think, text to deck feature.

And it basically creates a presentation from text to input. Yeah I also gave them walkthroughs with GPT and walkthroughs with Google Bard and showed them how to, find statistical evidence using Google Bard. It was just, it was really exciting for me. I'm a fan. I'm not against it.

It's still, I think, out to be seen what its best uses are. The output we get is not, I think, ideal or excellent. Certainly not even a replacement for anything right now. What are teachers doing to prepare for it? According to APnews. com, the New York City Education Department said that it's restricting access on school networks and devices because it's worried about negative impacts on students.

on student learning, as well as concerns regarding the safety and accuracy of content. But there's no stopping a student from accessing chat TPT from a personal phone or computer at home. Heard is just teachers trying to share ways of identifying when students are using that. There've maybe been a few conversations about people that found useful purposes for it, but mostly it's it's either.

Teachers grappling with identifying work that may be, I say plagiarized. So it's interesting that this sits in a gray area where the output of AI is considered the owned by the person who generated it, but it's not their work. And so it's still plagiarism. But they own the copyright to it, apparently, so it, certainly it's a gray area, and it has presented some issues, but. Different districts, K 12, and then for different institutions, everyone's trying to piece together something. At Victor Valley College, where my husband and I both work. I know we're putting together a committee on AI and hopefully we'll have some sort of point person haven't decided that title yet, but who will be there as a resource for instructors to go to when they have issues or if they're trying to revamp their curriculum and don't know how to do it in light of AI the other college where I work.

Pretty much people are ignoring it. It was presented as if, like, how can we catch them plagiarizing and stop? And even with the AI detection, I've seen it. I've tried a lot of different programs. It's undetectable. There are ways around it. And there's some red subreddit right now where they're figuring out, ways around the latest tracker.

Yeah. I just don't think it's realistic to frame it as plagiarism and something that needs to be avoided because you're just going to only penalize the students who aren't aware of the latest technology. Teachers are thinking of alternative ways to assess student understanding that may not be completely text based, right?

Or writing based, so that would be probably my first go to move to kinda combat it until we understand it and know how to utilize it, know how to check for plagiarism like you mentioned. So there's probably some alternatives to explore in that regard. How can you tell if something was generated by AI?

According to chat G P T itself, identifying an AI generated essay can be challenging, but a few clues might help look for unusual or uncommon language choices. Lack of personal writing style, inconsistent arguments or abrupt transitions, unfamiliar references or outdated information, and exceptionally fast or.

Violinist production. However, it's important to note that as AI models advance distinguishing between AI generated and human written essays can become more difficult. So just having immersed myself in the language that G P T generates, I think that is probably the easiest one to recognize. There's just certain structures that it always repeats in an essay.

, there's certain sentence structures you could recognize. Certain keywords that really aren't very common for humans to use that AI uses all the time. Crucial pops up, mitigate pops up, highlight, underscore. I know I use it way too much, but sentence structures are popping up again. And again, it's pretty much clear to me that either they used AI to generate their text or, they, they've fed in something and used AI to edit.

And I actually, I have a couple cases where 50 50, probably they did, but I'm not gonna say anything because I don't want to, I would rather have somebody get away with something than falsely accused because I think that can be pretty damaging to a relationship. Yeah, there's smoke, there's fire.

Yeah. And so when you're reading an essay that a kid turns in and you think, I've never heard this student use the word unequivocally, and I'm not sure that they know how to spell it. Maybe I would, and I see that in the, intro paragraph, I'd probably open up chat GPT and, say the exact same thing in chat GPT that the student might have, generate a.

Five paragraph essay about the causes of the Great Depression, for example, or whatever. And just look for some similarities. The biggest thing is knowing the student. At least in high school, we spend our lives with these kids. We are everything to them. They are teachers and counselors and social workers and so on and so forth.

So I know that student, I know what, how they write and what they say, and if suddenly it's more formal, or if it's not their normal vocabulary, or their way of writing, or even their cadence, or the length of sentences, all of these things come into factors like a style that student had, and if suddenly that style's different, it's noticeable, and I don't think it's, Difficult even to tell when a student's doing something, turning something that's not theirs.

How early are we in the stages of AI? So what we're looking at right now is actually generative AI. AI has been around for a long time in various forms. Most logarithms are basically AI.

We're not in the infant stages of this at all, where it's just starting to mature. Maybe we're like 21 years old and in the AI world, there's a lot of maturity to go. There's a lot of development to go. And maybe it's just never not a part of this really. It almost since the beginning of computers has been a part of it.

Will it ever gain consciousness and become something of threat to the humanity? No, I don't think so ever. I think we are just scratching the surface. We've heard of AI for a while now in, in science and technology and even pop culture and that sort of thing. But now that it's coming into the classroom we are, we're in the infancy of it in my opinion.

Which is the scary part, which is why we're talking about it today. But again, we, I think what will happen in the short term is it's going to be overly policed and controlled and blocked on in school buildings and things like that, and that's totally fine. But like I said earlier, it's not going to go away.

We need to investigate it. We need to find ways to embrace it and to use it. As a tool to help students ultimately because it's going to be a tool for them as they go through life. Yeah, I listen to a lot of podcasts about this and you have everything from this is going to usher in utopia to the robots are going to come and take over.

And I honestly, I don't know which one it is. I'm hoping it's the former. I think it's still very early. I remember seeing, I think it was a Star Trek movie way back when, and it was like that the future model of education was presented there and they were basically in these AI bubbles where they would have these experiences that were tailored to the student.

And I, I do think in 50 years, it'll be something like that. And that's so cool. How great is that? Yeah, I think it's really early on. I get the fears. I think there are some valid fears there, but I just don't think anybody knows. I'm optimistic. I could be totally wrong.

I don't know. What should someone know before they think about using ChatGPT for assignments? It's important to critically evaluate and fact check. Any information provided by ChatsGPT before including it in academic assignments. ChatsGPT should be used as a tool for generating ideas and insights, but it is still essential to consult reputable sources and exercise critical thinking.

In order to ensure the validity and quality of the final work. Definitely for GPT, especially it will give factually incorrect information very often, and it will sound, it'll present it in a way that sounds sophisticated and it sounds like it's right, but it's not when you actually check. So for instance, one way that I've found out that students were using GPT was in a works cited or a bibliography section you'll have a perfectly formatted works cited page, which is really hard for me to do anyway, perfectly formatted. But then when you actually check the links, they don't exist anymore because, they were generated from old information.

So definitely if you're using AI for any sort of evidence based writing, you've got to verify. So use it as a starting point, but then go back to your traditional methods and verify that information actually exists. You need to know is that you have to verify and source all that information. So one, if you're going to turn in any information, you've got to provide a source.

It's got to be a credible source, and it's got to be something that's academic generally. That's not Chad GPT because a lot of what sometimes you get stuff out that's actually completely off the wall and doesn't make sense or it's completely incorrect. So you have to be able to fact check that entire thing and know that it's all correct.

And if you have a source for it, you might as well just done the work yourself. So it becomes this, why even try? Same thing that you were told by teachers, the same thing that I was told by teachers, the same thing my parents was told by their teachers that, hey, taking a shortcut, cheating isn't going to help you.

You're sure you might get an a, you might get by this deadline or whatever, but in the long run, it's not healthy. And so there's only one way to, to do it the right way and do the work and put in the time. Sure. It's an awesome tool. You can maybe maybe you won't get caught.

Maybe you get a good grade. But in the long run it's not going to help you as much as you think it might. Thank you for watching TFD Deep Dives. For more episodes like this one, please remember to like, or subscribe. Reporting to you from Fairview Heights, Illinois, I am Ben from TFD Deep Dives.