# Exposing India’s Broken Education System After ChatGPT

## Метаданные

- **Канал:** Varun Mayya
- **YouTube:** https://www.youtube.com/watch?v=uLFMTJgMdNo
- **Дата:** 05.08.2025
- **Длительность:** 27:05
- **Просмотры:** 158,776

## Описание

In Indian classrooms today, students are using AI to write assignments they don’t understand, while teachers are using AI to grade those assignments. The result? An endless loop where one AI model is trying to beat another AI model, and nobody is learning anything real.

With this video, our aim was to talk to the students and teachers directly and understand how both sides are quietly relying on AI. We then get into why this cat-and-mouse game can’t last forever, and what it reveals about the bigger crisis in our education system.

We also explore why banning AI isn’t a solution, and how thoughtful integration could evolve learning. From personalised AI tutors to human teachers assigned the job of mentoring students how to use AI. And in a world where college brands often matter more than real skills, we then ask what students can do when the brand isn’t on their side. And what we found was that a laptop, AI tools, and self-learning might just be the answer. We then talk about how laptops powered by Intel® Core™ Ultra Processors could help bring personalised learning to everyone, and laid out 3 futures India could choose moving forward. Watch till the end to find out what those three futures are.

Learn more about Intel® Core™ Ultra Processors here : https://www.intel.com/content/www/us/en/products/details/processors/core-ultra.html

Apply for Senior Motion Graphics Designer Role at Varun Mayya - https://forms.gle/DhKxYneiurfGbV4P8

00:00 - Intro
01:04 - Students vs Teachers
05:32 - Chinese Room Experiment
07:22 - Why Not Embrace AI in Education?
09:21 - Why College Still Matters in the AI Era
12:17 - Learning Independently with Technology
17:18 - Essential Future Skills: Mastering AI Literacy & Prompt Engineering
19:13 - ChatGPT Better Than Teachers
23:17 - Three Possible Futures for Education

## Содержание

### [0:00](https://www.youtube.com/watch?v=uLFMTJgMdNo) Intro

So the other day I was talking to this teacher friend of mine and mid conversation he casually said that — yes I use AI to grade my students assignments. — Now I'd already known that kids were anyway using AI to complete their assignments but a teacher saying that he was using AI to grade it was something entirely new. All this seems like a giant hoax waiting to uncover. The student doesn't even write his assignments and the teachers don't even read it and somehow nobody has a clue. So like any rational person would do, I began my search for the truth. I wanted to find out exactly what's going on in colleges today. Because see AI is relatively a new concept and a college degree is 4 years long and many people today who are in college started their college journeys before the Gen AI boom. So I set out to talk to the students and teachers themselves. Ladies and gentlemen, today we're going to reveal the truth and the current state of our entire education system after AI.

### [1:04](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=64s) Students vs Teachers

— How do you do your homework these days? — I do my homework again mostly through AI. — My first step is to go open AI. So even though if you want to write it on paper, we charge you the answer and then write it. — Now if it was something like online based, I would just copy and paste and get it done with like I just see what's the homework, what the homework is, like what the questions are and then I put it into either chargity or some other AI tool and then I frame it myself. Usually I learn it from the answers it gives. — So we've been hearing a lot about AI use in colleges. Could you share where you have been using AI, how you've been using it? AI is now the go-to system which a lot of people are using. It's both students and teachers using this. — Being a scientist and educator, I've been using AI tools. — As an individual, if you ask me, I use it quite frequently. I mean, the time that we are in right now expects teachers uh to you know pick up the skills that you know we're actually teaching children. So that's how I or all of us got into using AI. — I kind of use it like an assistant which helps me to get things done quickly. So we use AI to build our lesson plans, jot out questions, assignment questions for students even with regard to remarks or even as simple as script that we need to prepare. — Do your teachers know you do this? — They do know that everyone is using air. — Yeah, definitely 100%. Some of them know, some of them don't know about this. — Do you think they know? You know, — yes, they know. — Uh yes, uh teachers definitely know this. They do recommend us not to use charge GBT but even they know we anyone's going to end up using charge so they'll just be like try to understand it as well when you're using charge and then write it down. — So it's like a open secret we know that they also know it — some of the students kind of are also surprised about whether the teachers are using AI. — Do you think teachers are using AI to grade your assignments? I would say not for grading the assignments but obviously to prepare for their lectures or for uh testing. Do you use it to grade assignments? — Yes, I do assignments. — Yes, we do. Absolutely we do. — They're likely to use it like 80% of the time they're likely to use it. — And why do I do that is because I'm going to take invest my time in one assignment and look line by line what they have written. It will take a long time for me. Instead of that if I use you know any generative AI which is there uh it helps me to grade give the feedback instantly. — What do your friends do? — Yeah most of my friends just copy it down plain exact however it was. — They use tragy just blindly copying it — and when exams come because we have offline exams that's where they can't do anything but in assignments they're getting full on full scores. So for example, if the deadline is tomorrow, they would sit 30 or 1 hour before that, go to the assignment page, copy the overall assignment and then go to AI and find whatever what could be the possible answer. — All internals are maxed out. Yeah, full on scores. It's 20 on 20. — But are they doing it cleverly? Are they humanizing it? — Yeah, that everyone's doing it cleverly. That's the thing. It was uh it's become too wide. They first check where are the plagiarisms are happening and then they humanize it through an AI or like just retype it. — Everyone's whatever they're copying they humanize it. There's tools like Quillbot all that you can even ask sometimes other AI to just humanize it a little and just uh instead of charg uh words that you wouldn't use like it's very advanced level of English instead of that just use simpler words and that would be humanized. So you tell it use simpler words. — Yeah, just use simpler words. That is more than enough to humanize it. — Do are you aware that they're using these humanizer tools? — Yeah. Yeah, I know that. — So, so how do you think they use these? — I'm interacting with student on a daily basis, right? I know their intellect level very well. When I ask question in a classroom, I know exactly what is the depth and suddenly I receive an assignment uh which sounds like I'm talking to an Einstein. I can make out that they have used AI. So, this is our education system today. The students use AI to write assignments and the teachers use AI to grade them. And somehow we're all fine with it. The students have no idea what the questions mean and the teachers quality of those assignments are. It's all being outsourced to AI where one model is trying to beat the other model. Think of it like the Chinese room

### [5:32](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=332s) Chinese Room Experiment

experiment. It's basically a thought experiment from 1980 developed by John Seal. So, back in the 80s, everyone was excited about computers. And if computers would be able to become intelligent like humans like if a machine passed the Turing test it must mean that it is intelligent. But John said no not really. Just because a machine could respond intelligently it doesn't mean that it is intelligent. There's a huge difference between acting like something versus being something. So he proposed this thought experiment called the Chinese room experiment. Let's say you're locked inside a room with nothing but a task. And the task is to basically respond to some questions written in Chinese. Every now and then, someone will slip the question paper under your door. But here's the thing. You don't know Chinese at all. Now, the first question is slipped through the door, and you start reading it, but you quickly learn you have no clue what the strange symbols mean. So, you look around in hopes of finding something that could help. And there it is, a thick rule book that basically tells you how to convert Chinese into English and vice versa. So, you open it and start converting the question into English. Then you figure out a reply and convert your answer from English to Chinese using the book and send it outside. Now imagine what it looks like from the outside. From the outside, it looks like you know and understand Chinese. But the truth is you don't. You're only following the instructions written in the rule book. Only pattern matching and feigning real understanding. Now apply that to how students are using AI. They're given some assignments and they don't understand. So they run it through their rule book, which is AI in this case, and respond intelligently. Now, from the outside, the teachers think, "Oh, the student is smart. He understands these concepts so well. " But all you're doing is running it through LLMs and feigning intelligence. There's no real learning. Just two people talking to each other and faking everything. That's the state of college and education we're in right now. So

### [7:22](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=442s) Why Not Embrace AI in Education?

let's ask another question here. If the entire work is being outsourced to Chad GPT anyway, why not just include it in our system? Why not stop pretending that AI isn't changing our lives in real time? We've had this entire idea of, oh, don't use a calculator because you're not going to carry it all the time. It used to be the narrative back in my day, but that's already failed. The entire idea of exams was built upon this broken belief that you're not going to have computers in your pocket, so you must memorize as much as you can. But look at today. It literally takes me 2 seconds to just pull out my phone and do anything I want. Not just the calculator. The entire stationary box has been replaced by the phone. In another example, I want you to look at CLY, which is a popular AI tool that helps you cheat on everything like your exams or on an interview. The only reason CL is becoming so popular and has raised $15 million from A6Z despite the fact that it encourages cheating is because banning technology is not sustainable. The CLY founder has literally said, "Today is cheating, but tomorrow it's fair. " Think about that for a second. As technology progresses, we're going to have more and more tools that let you do anything you want. We're going to have glasses with built-in LLMs that give you the answers in a beautiful hovering UI. We're going to have Chat GPT in our eyes plugged in to us all the time. Johnny IV and OpenAI are already building hardware products that let you do it. Soon, we're even going to have Neuralink chips available to us. Some of you might plant it in, not plant it in, but it's going to be inside your skulls and it's going to give us answers without even asking. It's going to happen at some point. It's already past the proof of concept stage. All this banning AI in college is not really sustainable. You can't ban something till eternity. Someday you will realize that you messed up big time, but by that time you would have already missed the train. Somebody has to wake up and shake the entire system into waking up as well. This can't go on forever. And the faster we adapt to this change and design our curriculums around AI, the better it is for us.

### [9:21](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=561s) Why College Still Matters in the AI Era

— So do you think college degrees will be useful though in 10 years? — I personally think yes, degrees are essential. College degree sets a curriculum for you to go through. It sets a path. So whenever you also program your NATSAs and GPS's to take you to a destination, it takes you along points. Whereas if you do it on your own, sometimes you're diverted. — Now even with all its flaws, it would be unfair for me to say that schools or colleges have no value. Some colleges do offer meaningful things, but probably not in the way most people think. See, college isn't just about learning new things. That part you can do on your own by just going to the internet and learning anything you want. It's been possible since preAI. So then what is the point of college? Well, among other things, one of the biggest values colleges offer is helping you get a job. In a way, colleges function kind of like catalysts and credibility boosters. Yes, there's some learning happening and there's definitely the practicals which I think is very valuable and there's also inhand experience and peer relationships. But if you went to a top tier college like an IIT or Harvard, it's much easier for you to get a job. That's because companies aren't always hiring based on actual intelligence, but perceived intelligence. They look at your CV and think, "This person got into one of the toughest colleges out there, so they must be smart. " That's because these colleges and institutions are already filtering for certain traits like intelligence, critical thinking, and hard work. How would you get through a J exam without those things? And those colleges then become credibility tags. That's the whole point of designing entrance exams so tough because then getting in becomes kind of a certification itself. You have already proved yourself going in. Now on the flip side, if your degree comes from a lesserknown college, it can work against you today. And it usually does. Even if you worked hard and learned a lot in your own, there's still that doubt in a recruiter's mind. Why didn't they make it into a better college? Now, we can debate whether this is fair or not, but that's how people's perceptions work and that's how the system has always seemed to work. You know, we learned this from A since A now has become a brand. I'm talking about the video editing school. If you're a video editor graduating from there, you don't face much difficulty today getting a job. It's like a stamp, right? And that now we have to be even more careful of making sure the people have skills because for that brand and credibility to continue, we have to make sure these people are good. Sometimes you even get placed without the interview process. That's because the very first few people who came out of the AV school were genuinely very good, which helped us build the reputation AV has. Now, a lot of the edte. So in their first batch they get the best students and they give them scholarships because they're like this is going to build the brand for the next few students because when these students go to companies or get placed in companies if the companies really like it they tell their friends hey I love this talent and they hire more from it. So a lot of edteexs have started doing this now anyway. So what

### [12:17](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=737s) Learning Independently with Technology

if you couldn't get into a nice college? What if you're just not good at giving exams so you went to a low tier college. So the brand is not working for you but you're still a smart individual. What's left for you now? How are you going to get hired when the brand is not working for you? Well, I have a solution and to be honest, it's very obvious. It's just learn by yourself. You have this god-like technology with you. And the thing is, we work very closely with laptop brands. And the truth is the reason for it. While we had options of so many different brands to work with, a laptop is something I truly believe in. Thing is, your laptop today is much more than a device that you can just watch movies on. It's literally one of the best features you can have. Think about it. You can go to YouTube and the internet and learn almost anything. And with the advanced AI models we have now, learning and education can get thousand times better. But why laptops exactly? You could probably do this on a phone, too. And like I've said, as a creator, there are so many categories of brands that want to work with us, but our default has always been, hm, we should definitely work with one PC or laptop brand. Why? Well, there are two reasons. First is laptops allow you to enter producer mode. And there's just something about laptop real estate, right? A bigger screen that lets you enter a producer mode. Think about how you use your phone. When you're on your phone, you're suddenly a consumer. You end up watching reals, scroll Twitter, or just randomly consume useless things. And you might not have spent too much time thinking about this, but ask yourself why you don't sit and have the behavior of scrolling reels on a laptop. It's strange, right? Like why don't you go through shots and reels on a laptop? It's not natural behavior. And this is a working theory is that when the screen size is larger and you're able to do multiple things have tabs, you enter this mental state where you're more mindful about things and think more like a creator. You start coding, learning and researching about different topics and go into rabbit holes. And this is true even if you started with gaming. Like for me, my entry into most of these rabbit holes was when I was young, I used to play a lot of Dota. And because I was on the same machine while looking for a game, I would alt tab and start trying other things. I would just play around with things. And it only makes sense, right? Because since the dawn of the app store, all the social media apps have been optimized for smartphones, not for laptops. And over time, your brain starts to associate your phone with social media apps. Over time, your brain starts to make social media synonymous with your phone. It's basic psychology, which is why laptops are so important, right? I' I've also written this in my book, Pajama Profit, that if you give me a laptop and an internet connection, I'll make my money anywhere in the world. And out of the many laptop brands in the world, I really love working with Intel. And there's a solid reason for it. The reason we keep saying in our videos, Intel, Intel, is I have a very deep connection with them, which I told them as well when I started working with them. When I was growing up, the first machine I ever used was an Intel machine. The first code I ever wrote was an Intel PC. That's where it all started. And I promised myself that when I got older, I'd make it full circle, right? I didn't know it would happen in this way. That PC made it possible for me to teach myself and build stuff. Yes, Intel builds fast and powerful machines, but there's this deep appreciation for Intel because I grew up on it. I still remember playing the old Mortal Kombat games on it. I still remember doing everything I did on it, and it makes me want to work with them. They've always been a part of my journey in some way or another. And who knows, if I didn't have that Intel PC growing up, I might probably never have been an entrepreneur or started AOS. The second reason I love working with them is because of the advancements they've made making AI more accessible to kids in India. Their latest machines aren't normal PCs. They're now what the marketing world calls AI PCs, but they're for the right reasons. These machines are powered by Intel Core Ultra processors that can run local models and give students a hands-on way to build and learn. And they've got all the bells and whistles, the NPUs inside. And to be honest, they make AI accessible to even the remotest locations of India. AI today comes packaged with a device, even to people who've never heard of Chat GPD. And yes, there are people in corners of India who've not heard of it yet. Every student today can experiment with LLMs and learn about AI no matter what city they come from. No matter if they live in high or low connectivity areas. Now that's a big deal because only about 60% of rural India has internet access. And even where connectivity exists, remember that only 55% of schools have digital access versus 27% in rural. You may be the exception and have had a good computer lab in your school. But for most people who have computer labs in their school, those computers are very outdated. So think about where they're experimenting. They're experimenting on a very outdated machine. And that's why I believe everyone should have their own AIPC. And I think that whether it's running a chatbot trained on your textbook, editing a school video project, or just experimenting with AI and learning new things, there is some correlation I've seen with my life on people who had a laptop for a very long time versus kids who've grown up with just a phone. And I think it's definitely more causation than correlation. That's why I'm very vocal about laptops, whether Intel or not. And I love working with Intel. But whether Intel or not, I think everyone should own a laptop because it allows you to enter producer mode. be more of a creator than a consumer.

### [17:18](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=1038s) Essential Future Skills: Mastering AI Literacy & Prompt Engineering

But the idea is I think we're entering a world where you can just learn anything you want and open yourself up to future opportunities without relying on anyone. And the best antidote to sitting brainlessly and scrolling on your phone, which a lot of people are now doing, is to use a device that doesn't enable that easily. And I think that you can actually sit down and future proof yourself by learning critical skills like even prompt engineering. Right? There's so many people that laughed at prompt engineering. But it's actually a very valuable skill thinking in systems and AI literacy. One of the kids that we got said that hey I think something new needs to come about called AI literacy which is how good are you using AI? And we have another concept called computer literacy but we don't really talk about another concept called phone literacy because it doesn't matter or mean as much. And I think that will matter in the post AI world. So I believe that no matter what is happening between students and teachers now in this world as long as you have your own device, you should just learn by yourself. You can learn how to code. contribute to open source projects. YouTube is free and open to everybody. And I think you should build your CV that way because companies still care about end outputs and end project. Trust me, when recruiters hire you, you're going to have a better CV than many of your peers. You can participate in global hackathons like NASA has NASA's space apps challenge or you can participate in Google code jam and you can build real world solutions and I think that's the solution to this assignments are sort of a proxy for oh maybe this person knows something but building things when you put something out there that you've worked on and that you've built and you've put in the real world I think that is the real measure of how recruiters will hire you like look at levels IO right he built an entire game from scratch with nothing but his computer AI and an internet connection and that's the idea it is truly democratized intelligence and I think now everyone has access to the intelligence AI doesn't tire it can answer your questions back to back and it doesn't make sense that we ban technology like this do you feel like there's ever going to be a

### [19:13](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=1153s) ChatGPT Better Than Teachers

point as AI keeps getting better where it might replace the need for teachers — uh no I don't think so AI can give you answers but AI can't tell you that uh you know I really believe in you can do it much better So the coaching, the mentorship, right, — the — motivation, um I think in education, this discipline and ethics is very important, the values which we inculcate. — Now I know it might kind of seem unusual, but the truth is that AI can be as good as your teacher when you know how to use it. And the numbers back it up. Back in 2022 when GPD 3. 5 was around, it could take major exams, but it wasn't doing great. It scored about 82% on the SAT, 40% on the LSAT, and only 10% on the bar exam. Basically, it was just getting by. But then GPD4 came in and crushed all of those exams. It scored 94% on the SAT, 88% on the LSAT, and 90% on the bar. And it's not just one metric. While GPD4 scored about 85%, far better than most medical students, GP 3. 5 only got 58%. Now consider how much your college teaching has evolved in the last couple of years. Probably not much. Yes, there are many dedicated teachers and professors doing great work, but collectively the pace of improvement just doesn't compare to how fast AI is moving. But human teachers are still critical. They give you guidance, mentorship, and sometimes peer-to-peer life advice that AI would have never have been through. But if your son or your daughter has already missed the chance of getting into a great college and are left here wondering where they're going with their life and career now turning to AI and learning by yourself is useful. And the data that we showed you speaks for itself. Plus, it's just a scale problem. There are not enough good teachers for every student. Walk into a college and you'll find one professor managing 100 plus students. Despite Benjamin Bloom proving decades ago that students who get one-on-one tutoring perform two whole standard deviations better than the ones in a normal classroom, it's the difference between 50th percentile and 98 percentile. It's like jumping from being average to top 2%. We are still stuck with one teacher per 100 students. It's unbelievable. With AI and technology getting better, I think every student can have a personal tutor that doesn't just give them answers, but actually guides them into solving complex problems. and think about what this can do for students. So you might think this is a little disappointing which is that Von you're saying that the college system right now or the school system right now is two sets of people that are faking assignments with each other and you're saying that the devices that we're starting to use from the laptop to the glasses are going to get better and you're eventually going to be able to answer all these questions and right now it's viewed as cheating but over time it's going to be viewed as oh this is normal then why even go to college school and as we've explained the value is in all the other stuff the stuff that you know making friends, learning how to speak, learning how to be reliable. I think there's a lot of value in the soft skills and I think as a nation we have made it too much about hard skills and I think the hard skills are obviously very valuable but now can be self-arning and even though I've gone on record multiple times to say that everyone should own a personal device and learn by themselves and make it a habit to learn by yourself and experiment by yourself and I do believe by the way just because Intel is a you know partner it doesn't mean that I would say something I don't believe in and I think that laptops truly are thevices device like I don't know why this is the case which is I'll pull out a phone and then I will you know eventually start scrolling reels or going to messages or looking my DMs but on a computer whether it be desktop or laptop I'm really inspired right like I sit down and I'm like today you know I could download this new tool like last week I was playing with a new tool called Gaia which is a tool used to make landscape and terrain in Unreal Engine and I was just playing with it I know I don't need to be an expert with it right now but I want to play with it to learn to see what's new and I think that really matters and it's something where your teachers are simply not trying all the new tools that are coming out in whatever space that you're in. And I think especially with stuff like AI, you have to experiment by yourself. So what

### [23:17](https://www.youtube.com/watch?v=uLFMTJgMdNo&t=1397s) Three Possible Futures for Education

now? I've run through some of the permutations and combinations and I've arrived at the conclusion that India has three potential futures moving forward. Future one is the one we're already slipping into where all of this just continues. We build more detection tools and start banning AI use in colleges. It won't work, however, because again, students will get better at hiding it. And as you can see, they already have a terminology now called humanizing where they put the AI outputs through something else to make it sound like it's not come from AI. Future 2 is where the entire system collapses. Companies start abandoning degrees and start hiring based on their skills alone. Similar to what's happening in college, it's also happening with CVS and automated applicant tracking systems. The CVs, well, they're all generated with AI and the applicant tracking systems are using AI to figure out which candidates to actually interview. So, companies will eventually start creating their own tests, their own assignments, and stop trusting colleges altogether. As you heard, colleges are now giving 20 on 20 in every assignment. And that's not sustainable. Traditional universities slowly start becoming irrelevant and fade out. See, people will still learn because there's AI now, but we lose the mentorship, the peer learning, the human touch that college provides. It all becomes just you and a screen and nothing else. The third future, the one future that actually makes sense. We use AI to give every student a personalized tutor, just like Bloom imagined. Teachers then focus on what AI can't do, like mentoring, guiding, and helping students. They help them how to think. AI use becomes factored in and every test will and should assume that students will always use AI. So tests become more about creativity and problem solving and less about memorization. Yes, the degree still exists, but they start signaling something real again. The solution isn't to outright ban AI in schools or colleges. It's simply not sustainable. And it goes back to the calculator example I gave you. We stop pretending that AI is not taking over and start preparing for it. This means there's only two kinds of reality unfolding as we speak. One in which schools ignore AI and keep running the system with their outdated textbooks. And another is where students learn in isolation sitting at home with their devices using AI to teach themselves everything except the real world experiences all alone. So my opinion is that the best part forward isn't choosing one over the other. It's kind of merging the two. It's about bringing AI into our education system and designing our curriculum around this new technology so teachers can become guides and students can get personalized support and the whole system evolves instead of simply getting replaced. We need to stop pretending that AI use is bad or somehow unethical and start building systems that not just teach but also grade people on how good they are at using AI. Imagine the amount of energy it saves, right? that students aren't wasting uh their time hiding their AIUS or even being penalized for it and the baseline gets evened out. So anyway, that's what I think. But as you all know, these aren't simple problems. I'm sure in the comment section, you can spend the time and energy thoughtfully telling us what you think the future of something like this should be, what the future of the education system can be. Because AI is now that shift where we have to make a decision stop complaining about the education system and say look this thing now will break the education system. How do we fix it? No one has the answers but your comment can guide everyone watching what the future could potentially be. That's it for me. Hope you like this video. Bye.

---
*Источник: https://ekstraktznaniy.ru/video/11969*