Subscribe
Share
Share
Embed
The Honor Roll School official podcast: home of the school's up-to-date information adding another layer to our school communication --- This podcast is for informational and entertainment purposes only. We make no representations or warranties regarding the accuracy, completeness, or applicability of the content. SEG Inc. expressly disclaims any and all liability or responsibility for any direct, indirect, incidental, special, consequential, or other damages arising out of any individual’s use of, reference to, reliance on, or inability to use, this podcast or the information presented in this podcast. This podcast and all its content, including but not limited to audio recordings, show notes, artwork, and branding elements, are protected by copyright laws. All rights reserved. No part of this podcast may be reproduced, distributed, or transmitted without prior written permission from SEG Inc.
Hello. This is Mrs. Quintero, the proud Head of School, of the honor roll school, and welcome back to the pod the dolphin podcast. Please follow us on Spotify and Apple podcasts. Search us as the tank thrs official dolphin podcast. Don't forget to turn on those notifications so you don't miss the next episode. And for this week's podcast, I had the opportunity to catch up with senior student Rishi Gupta from one of our spring education sister schools at basis Independent School in Fremont, California. We had a great conversation about his experience with AI as a student.
But first, let's dive right into what's coming up.
Reminder that the fall school wide survey was launched this week in batches through the first week of November. It is so important that all families take a few minutes to complete. The more responses, the more accurate the results. We look forward to the results to create our school improvement action plans based on your valuable feedback. Partner sponsored pumpkin patch will be on October 23 at our Riverstone campus and on October 24 at the Sweetwater campus. Second and Third grade Halloween concert will be on Friday October 24 as well as the Middle School Halloween social that evening, prospective family open house is on Saturday, October 25 starting at 9am at both Riverstone and Sweetwater campuses. Parents, if you have friends and neighbors you know would love your school community, please let them know. And now for my chat with Rishi.
AI is really so exciting. It really is. And you know, it all comes with all the good and all the challenges you know that come with using AI in a school setting. So Rishi just recently presented to the honor roll school teachers, because, you know, I really felt that it was important for us to understand how we're going to evaluate having AI in our classrooms, and then we have to keep reevaluating, because AI is always changing. It's changing daily, and how are we going to go about using it in classrooms, like AI is inevitable, right? It's inevitable. So I think we have to embrace it, and then, you know, really understand it, and how do we get this across to our students? So having heard Rishi, and having heard, you know, his really great presentation, and having a student's point of view, I think is important for us to have this conversation, you know, for our parents to hear, you know, what their students are thinking and how it's going to work out for them. So take it away. Rishi, you go, and I'm going to insert some questions every once in a while, because I'm, I'm not versed on everything, but you are,
yeah, definitely. So I think that what really kind of started this project was with my personal experience using AI. AI has a wide variety of functions and features, and when I used it to help me with coding, I noticed, you know, there were some great things about it, and there were also many flaws. So when I looked at AI in education context, I think my biggest thing was building trust. I think in education, the biggest and most important thing is the relationship between students and teachers, and AI is a tool that, if it's misused, can easily break this relationship. So one of the things that I've emphasized is building the trust between students and teachers. And the other thing which I've emphasized is that not all AI is good, but not all AI is also bad. AI is kind of in this gray room where there are some benefits and there are some things that are really harmful. So we really need to properly address it and get into the specifics of AI, rather than thinking about as like a black or white picture of AI is good or AI is bad. Definitely has multiple features that as a student, I think that we can definitely implement in an educational context. It has other features that I think should definitely be left out in an educational context. But overall, we really need to start getting into the weeds of how we look into AI as a whole. Your perspective as a student really is important. Like, you're just on the precipice of, like, all that, all that can happen with AI and all the good that comes with it. And, you know, I think too the teachers sit and go, Wow, these kids really know so much I've got to keep up with this. And how do I keep ahead of the game and make sure that they're learning? I think when you present it to the teachers, one of the things that I wrote down, that I that I just love, is that you.
Said, you know, for students, like, if you remove the AI tools, could students still complete the task? A lot of problems, which I've seen with AI, especially in a coding sense, is a lot of people will plug in, like a prompt to GPT, and they'll say, Hey, can you write this really large and extensive code, and then they write the code, they get it, and then there's bugs with the code, and they literally don't know where to start to fix it, because they don't have that coding knowledge. And they're so reliant on AI, it's as if they can't do anything else without the usage of AI. And I think that this is a really big problem, because we can't be so reliant on this one tool that we can't even verify how accurate it is. I mean, AI is still far from being perfect. It definitely is great. I mean, it provides a lot of features that have fascinated myself as well, but there are still a lot of biases with the data it's trained on there's still a lot of misinformation, and sometimes, most of the times, actually, it hallucinates when you ask it a question, and it can give you responses that just aren't accurate. So I think that while it's great as a tool, it should not be anything to replace a student's skill. And that's when I presented to the teachers I had mentioned that I view AI like a calculator. Calculator is like a tool which helps and aids students or anyone in math, but people don't know how to do the math. It's not like they don't know how to do addition subtraction. It's just that the calculator aids in making that process faster and more efficient. Yeah, yeah. And to what you were talking about, I think you had a really cool slide up about what are the students should use and what the students should not use. So can you talk a little more about that for the parents? Because I think that was a very great slide. I think a lot of the teachers are doing a screenshot of that one.
Yeah, yeah. And to what you were talking about, I think you had a really cool slide up about what are the students should use and what the students should not use. So can you talk a little more about that for the parents? Because I think that was a very great slide. I think a lot of the teachers are doing a screenshot of that one.
I'm personally not the neatest note taker. My notes are all over the place, and AI can really help with simplifying the process of taking my raw notes and my unstructured, messy notes, and you can really take that and form it into a legible format that's easy for me to learn and comprehend, so when I have to review for tests, I don't have to spend time actually going through all the notes and fixing them. Rather just ask the AI to help me with that. Another example is my history teacher last year, he had given us some questions, and he basically told us, okay, here's some questions to help you, to guide you on the specific topic, you have to read these pages of the textbook. Now, a lot of people, they got lazy, they didn't want to read the textbook. And what seems obvious enough at first is from a student perspective, just plug in the questions to AI. Just see what the AI has to offer, and that's your study. But I realized that AI wouldn't be super accurate, especially with many of these questions, so I took an alternative approach, where I still plugged it into the AI, and I asked the AI to just give me some bullet points regarding the questions. But then I asked the AI to or not, not as the AI. But what I did after that was I had read the textbook myself, and as I went through the pages, and as I read through the content, I looked at the notes that the AI generated, and my learning was by fixing the notes and actually either adding more information or fixing the AI's mistakes. And trust me, there were a lot of mistakes that I had to go through, and I felt like that was one of the best ways of genuinely learning a concept, because it wasn't just I'm reading a textbook. I'm just writing down some words on a piece of paper, and there's my notes. It was genuinely me going through each and every line, each and every word, making sure everything is accurate, and actually changing that in my notes to reflect if I've actually learned and understood the concept or not, and that really helped me get a really strong grasp of the concepts that we were learning. And for me, last year, history was definitely not a struggle at all because of the way that I implemented AI in my study. More broadly, I think that AI can be used as a tool for further exploration, to allow students, even teachers, to be creative in their use case, in their uses of AI. For me, I had given the example of my own personal lab, where we were given a task which was relatively straightforward, but there was still something which I wasn't satisfied with, and I wanted to really further explore it and get a better understanding of what I was dealing with. So I used AI to help me make certain 3d graphs. And the graphs didn't turn out perfect at first, because AI is definitely going to make mistakes. It takes time for the AI. To learn, and AI is you can treat it like a human. You can't expect humans to just instantly answer a question. You really got to help them walk them through the process. And AI works pretty much the same way. So from the initial graph study gave I'd worked with the AI, and I kind of talked to it about, okay, we need to fix these parts, or these parts can be different or and I kind of showed him, why is the graph looking like this? Or why is this a feature? And slowly but surely, we were able to form much better graphs. And on top of that, I was able to really answer my questions that I had and I was just simply curious about, and they weren't a part of the lab, but with the help of AI, I was able to answer and get a better understanding. So I think that AI in these contexts can definitely be very useful for student learning. I think there is an element of students using it responsibly that's always going to be a feature with AI, but I think that there are students that definitely do use it responsibly and do have that capability. So we really want to open up to students exploring AI, because if we don't, and then they suddenly get into the industry, and suddenly people are using AI and they don't know how to use it properly, and they're just going to take it for granted and fall short to its biases and all of that that it's really going to be bad for them in the long term.
Where do you see kids not using it right? Where do you see things that you know they're going to want to try but they really shouldn't like? From your point of view,
it's a trade off. It's how much, it's kind of a moral question of, how much do I want to sacrifice using AI for like, an assignment or a homework at the cost of the moral so I think that in a lot of like high pressure standards, when students maybe have a lot of work or they're kind of just lazy, they just want to get something done, they think that they can just use AI and simply, you know, bypass everything, just quickly finish the assignment. And, yeah, I mean, that's it's done then. But I think that the point of homeworks and assignments are for students to get a better understanding of what they're learning, to do more practice on these concepts outside of the classroom, and I think that using AI is more of like a short term solution. It's almost as if, like, the AI is, like, generating code for you. It's like, you'll think today, and you'll be like, Whoa, I'm generating code, but you won't know any code yourself, and that's going to be a problem in the long term, and that's actually what's a huge problem, which is ongoing in industry right now, is that a lot of people walk into computer science and they say, I know how to code. And the moment you remove the AI away, they do not know anything at all. And it's become a very common trend, and that's one of the things that now, when people are hiring computer scientists, they're being a bit more cautious about whether these people genuinely know the topic, or are they just using AI and kind of just faking their way in.
You know, the other day I had, and I know you brought this up with our teachers, I think one of them had asked you a question about art in AI, right? Using AI for art. But I did have a couple kids come up and like, Oh, Miss Quintero. Can I Can I please take out my phone so I can show you something? And I'm like, okay, so they take it out and they open up just a photo of this like lion coming through down the street right, going after a little child. And I was like, Wait, hold on a second. That's not real. And they're like, Yeah, but just look how it looks. It just looks so real. You know, I think you had a good message about looking for the fakes, right? What are the fakes? So give me a little little talk about that.
Yeah, I think AI generated content in terms of like videos is becoming a lot more prominent. In fact, just to quickly add, there's a new AI, I think it's called Sora AI. It's a complete like video type AI. It will generate all like, actual like video for you to post on like social media or whatnot. And it is actually relatively accurate. It's still very detectable that it's an AI, but if people have been worried that it can get it's pretty scarily good. I think it's a Google's AI, but it's pretty scarily good. But yeah, in terms of just detecting, like fakes, I think in terms of videos, it's one of the more simpler things, because the one feature of humans that AIS will never have is that we humans are imperfect. We have many irregularities. When we walk, we don't walk perfectly. When we dress, we don't dress perfectly, maybe of a strand of. Hair too, which just sticks out, or maybe your clothes are slightly crumpled, and that really distinguishes us from Ai, because AI is kind of this perfect view of who we imagine ourselves to be. So when we think about AI and AI videos, anything that seems just too perfect, like maybe it's like a cooking video, and there's like a cut into like a certain fruit, and the cut is nearly perfect. You know, that's really high signs of it being AI. But, yeah, it's getting more and more interesting to see how we can really distinguish AI from real I think that in terms of these AI companies and these features that actually generate these videos, I think it's gonna be really important that they actually put watermarks and they actually label it as AI, because the moment you don't label it as AI, and this was a criticism of the new Sora AI, its watermark is kind of Like at the bottom or maybe to the side, but it's not covering the entire screen. And people were saying that it could easily be cropped out, and then you just have the video, which has pretty much everything that you would want, without the watermark. And it's getting a bit more worrisome, because if you have no way. At some point there's going to be really hard for us to detect whether it's AI or not. And if we don't have a way from these actual tech companies to distinguish or to actually mark a video as AI, it's going to cause a lot of problems with misinformation and bias and all of that, obviously, on social media, things can get very complicated and very heated, very fast. Misinformation, especially in like, a political sense, can spread insanely fast as well, and these videos are definitely going to contribute to that.
Yeah, so Rishi, I'm gonna tell you. Like, again, AI is inevitable. Like, I think as students and teachers and schools that we have to, you know, navigate it. I think have discussions about it. I know you're in high school, and it's just a different way of looking at it for you. And, you know, we go through middle school here at the honor roll school. So, you know, I think dipping down into into middle school and and really address, not addressing it, but just expose, exposure, right? And discussions, I think are really important
in terms of usage of AI. I think a lot of people have different ways of like approaching these chat bots. And one thing I'd really strongly advise everyone to learn is how to actually prompt an AI to complete a task. Most of us will probably write a sentence like, can you just give me 10 bullet points on a certain topic, or can you give me a brief summary of this? But if you really want the you want to limit the AI and its biases, because if we understand how AI actually learns, it's learned on billions and trillions and even more amounts of data. And when it answers your question, it tries to just pull out data from these various spots. And sometimes it can get the context wrong, like, maybe ask it a math problem, but it accidentally gets a reference from an English problem because of the way that you entered it. So an example of like, proper prompting would be, like, you start off by saying, imagine you are a math student in a specific math class, and you're in maybe, like a test. It's like high stakes or something. And then you give the question because it prepares the AI to actually solve the question, rather than just being any question out of the blue. It's almost as if you asked a person like your best friend, or, you know, anyone, like a teacher, a student, a question. If you just tell them the question straightforwardly, they may not realize the context of the question, but they may not get it at first, but if you tell them, hey, I have a math question for you. This is on my math test or my math homework. It's problem five, and I really need help with this, if you kind of give that context, it really helps the person understand what specifically you need help with, and where they can truly assist in your learning of that particular problem. So I think that AI prompting is something that a lot of people don't really think of, especially with usage of chatbots, and something which I strongly advise anyone listening to this to actually go online and really learn how to properly use an AI Chatbot. Yeah, and the more we talk about this, the more we're going to learn from one another, and the quicker we're going to find a solution, especially in education.
Remember, you can find the pod the dolphin podcast on the weekly dolphin days newsletter, on the thrs app, and when searching us on Spotify and Apple podcast, look for the fighting dolphin logo.
This podcast is for informational and entertainment purposes only. This podcast and all its content are protected by copyright laws, All Rights Reserved. No part of this podcast May. Be reproduced, distributed or transmitted without prior written permission from S, E, G, Inc, you.