Why This Indian AI Refuses to Give Students the Answer

0:00 Now, the easiest thing an AI can say to a student is great idea costs nothing and everyone goes home happy. Well, Udium's AI mentor is built to say something else. A student in a government school sends a business idea on WhatsApp and instead of applause, she gets questions. Who will buy this? What will it cost? And well, what would you change? Now the mentor is called Sari. says it is now used by more than 7 lakh students and 25,000 teachers across six states. Mr. Mechin Maheshwari is the founder and CEO of Udium Learning Foundation. He's an engineer who started at Yahoo, joined Flipkart when it ran from just a two room house, built its technology team and later ran its people function. He is one of the founders of Plaksha University. Udyam, a nonprofit began in 2017. Mr. Mechin Maheshwari joins me exclusively on front page to well discuss what I believe is going to be a fascinating discussion. How are you doing me?

1:14 >> I'm well. thank you so much Sudi for having me on your show. >> Truly a pleasure. truly truly a pleasure and it's it's fascinating work that that you have initiated and it's truly impacting the lives of so many families. I wouldn't just say students, I think families in a big way. So kudos as far as that is concerned and I'd actually like to start there since I just mentioned the fact that you know engineer you built an entire team at Flipkart etc and things like that. What then kind of inspired you Megan to kind of move in this direction?

1:50 >> I think Flipkart's success and what that meant financially for me and my family allowed me to really think about key who am I? what do I really care about? And as a middle-ass boy who' just gone by goals that were shared, get into college, get a good job, get married, have kids, and you'd done all of that and done it well and succeeded at them, but never really paused to think right and it's not like I didn't enjoy any of this. I loved technology. I loved building Flipkart.

2:29 but it this introspection journey of about a year and a half which included bunch of experiments with vibashna with yoga etc allowed me to figure I get joy out of people succeeding. >> Nice. >> Right. Like the emotion that probably best describes a large part of me is the sports movie emotion. When you watch that 1983 and everybody's laughing you you have no chance. Right. I could still go and win. That to me is the story of humanity.

3:01 >> Wonderful. Wonderful. >> So that's that in some sense was the inspiration. There is a personal story. when I was 17 like most kids I'd worked really hard to get into IIT >> and I failed to West Bengal like the number of failures I faced and it felt like the system was willing to give up on me you're good for nothing >> and that was very very sad those were some of the saddest days I had get into a good college and you couldn't do it. You feel like and I think as I went on to do engineering from previous and then joined Yahoo and had a good career but that feeling didn't leave the system is willing to give up on actually 90 plus% of its kids.

3:56 >> Correct. >> And that's because >> yeah because they don't get into because we have only that many seats. >> Correct. >> Right. Right now we are fighting about how well are we conducting an entrance exam >> but the reality is that hey only one or two out of a thousand get in right across entrance exams. >> That's true >> and it's not like people have immense potential. So that in some sense is what led to Udam being born as there is a lot of human potential especially there is a lot of youth potential and our current systems don't necessarily find ways >> of letting young people shine of helping their potential come out. we have fixed pathways we select >> if not we'll reject you and it's like a manufacturing unit that rejects 95% of its produce.

4:48 >> True. And and and I think you've touched upon the right chord in in a big way, Meckin, because I mean you yourself and I've I'm so sorry that you had to kind of go through that weirdly dark time when when you were about 17 18 because I mean psychologically it can have such a huge impact and and like you very well very well mentioned the fact that it's not like people don't have potential. People have fabulous potential. Look at yourself. I mean you go on to build Flipkart which is one of the biggest brands probably globally and the fact that this was the same system which probably made you wonder about your own potential and understanding. Is that then the reason that you chose entrepreneurship as far as Udyam or Sarti was concerned visav for example something like a math or a reading curriculum was that then the motivation behind it?

5:43 yes it's broader in my mind I think I'll try and see if I can connect a few dots right so first is u the limitation of the number of opportunities >> like our education systems number of opportunities in higher education and then eventually in a in a job in a good stable job is fairly limited less than 10% of India's workforce has formal jobs >> sure >> right so the remaining people do not have formal jobs So only a few will get into a good job and our education system only celebrates when you get a 1 cr package on day one from camp >> correct >> don't so the reality is that system is designed to reject 90 90% kids so we are a selection or such but this is at the end of the day unfortunately the burden of colonial rule that that one is still kind of you know crumbling under >> because the design the system was designed to select is officers who could govern the country right and >> it remains the largest aspiration of youth in India that correct >> right but there is immense potential in the other 90% and we haven't found ways of letting that emerge so entrepreneurship offers an agentic way what do you want to see in the world what's your potential and unlike approaches is that look at for example vocational education or skilling which are almost determined aspect like you want to be a creator you want to be a YouTuber you want to be a dancer you want to be like create a 3D printing company please go ahead and do that right >> how can the system enable you to do that is a is Udyam's effort >> very quickly ly I' I'd like to kind of dig a little deeper on this with you when when you just mentioned regarding you know the numbers that you you shared which unfortunately are abysmal and very depressing in terms of you know formal formal jobs etc and things like that you know we keep talking about I mean you yourself are an engineer me we talk about the fact that you know how India is somewhere more of a a champion when it comes to engineering talent etc things like that and we can you know go ahead and beat the world on it. We did it with the IT services perspective for example all of that but there are very still according to me there are huge yawning gaps in terms of that talent truly emerging to make a thumping sort of a difference. Do you agree with me on that? I do. I think there are a handful of great colleges and institutions right and even in engineering I feel like having the IITs the NITs a reason to in some sense try and create PLA is >> that hey technical education can be reinvented can be rethought it need not be as theoretical young people can get far more experience of doing things of building things in a world where technology is so easily accessible and available with open source systems etc.

9:08 As in build stuff, build projects, build things that are usable >> even while you're studying. So helping young people learn the art and capability of building things that are useful while learning the theory and research underneath it I feel has been a is an important shift to try and make and a university like PLA is trying to do that. most of us most of our education system and I'll tell you a very very sad story very recently I was in a or at a UP secondary education school >> right or vocational education subject typing >> imagine typing is still taught as a vocational truly tragic >> I learned typing in 1995 asdfract and exam question as in basically down quity in the right order and don't make a mistake right like that is what we are teaching our kids that's how ancient theoretical unrelated to the real world and these same kids are able to create reals and create presentations that are far superior, right? So, so the system doesn't keep pace with with the world and with technology. There are no ways of updating the system that hey what should young people today be learning? Young people today are learning exactly the same physics, chemistry, maths that I learned 30 years back.

10:56 The irony of it all to be honest you know when when when one tends to talk to people from let's say you know all walks of life for example leaders at tech companies etc things like that and when when I've asked them this regarding the fact that you know what are you looking at as far as when you're hiring somebody from India etc things like that and I've pointedly asked do you look at leadership leadership skills. They said, "Yes, we do, etc." But the example you just gave, your leadership skills, sir, >> you're not letting the mind fly.

11:34 >> Yeah. Yeah. Yeah. No, I think our system and we touched upon it already, so I'll build more on it, that our system was designed for creating followers, for creating people who were disciplined, >> correct? >> Who could execute the command coming from abroad, >> right? not question it not contextualize it not figure what is the right way of doing it in this condition so sal pucha is not necessarily seen as a positive >> it's almost frowned upon sir >> exactly it's disrespect in a class not standing in line not being quiet like we we dull young people out of creativity and curiosity into straight lines and followership take away their curiosity and questioning or imagination and tell them this is all there is and I think a little bit of unleing of what young people can do if we were to just stop and ask that hey what can young people do you will see magic >> right correct >> so very early on when I was starting with them we came up with three core learning principles which have become pedagogical principles out of them and a fourth has been added first was learning by doing.

12:51 It has to become more doing. Second was it has to be in the real world. It can't be the simulations or imagine it is actually in the real world. Get out on the street. Get out in your community and do. >> And finally and most importantly, it has to be it has to be with agency. Right? Funny that agents have now become more AI than human. But when I created these three principles back in 2016, 17 agents and can humans be agents, can young people be given the agency.

13:31 Sure. Right. Sure. >> We finally added a fourth a little later which is outside of your comfort zone. Correct. >> Right. That you already know how to do stuff. you already know how to make a say a great sari right sell it right add more to it do something beyond what you already know >> which is what stretches your learning or pushes your learning so those are the four core learning principles and I see a good entrepreneur >> doing all of them >> right that entrepreneurship is all about doing that are deeply connected to the real world around them they are observing they're seeing opportunity there is nobody to tell you what to do so you have to make the choices. So all that agency is constantly there and a good entrepreneur is constantly pushing the limit on right. So those four principles which are which apply to entrepreneurship equally apply to learning and my effort has been to enable learning to happen in a way that entrepreneurship happens >> and it's actually showing up in the numbers as well. Let me just share this and correct me if I'm wrong Mecc if the numbers are wrong but novel ideas rose 160%.

14:45 And articulation 125%. So wonderful in terms of those four principles definitely are working. I want to understand this from you now from let's say the outsiders lens who scored them. That's the first thing that probably pops up in my mind. And also I want to understand was it probably the same AI that does the coaching or is there a distinction? >> Of course there's a distinction. So quickly just let me explain the numbers first and then answer the question that and before I answer the just the context behind it that when you are in a government school and you're asked for an idea right >> what happens most of the time is that you're shy you don't want to give out ideas because there'll be judgment like we are a reasonably judgmental society teachers so I think just shifting from sharing publicly to sharing on a WhatsApp chatbot >> right is between you and the chatbot and the chatbot does it very non-judgmentally that itself leads to a huge jump in number of submissions of ideas >> point so that's number one two is what we've seen and I'll give you an example that when students think of ideas there is a full orchestra going on in their head >> but when you ask them to communicate it Right. Again, we as a society haven't enabled communication, haven't taught communication. Right? They said, >> right? So, I'll give you an example. The student has thought of a brilliant idea.

16:27 And when we ask them that, hey, what's the idea? They're like, >> okay, that's a it's a noble idea. It's ide right. So the guy then gently asks you that okay how will you prevent stubble burning? Then the student says oh I'll buy the stubble from the farmers. Okay great you will buy the stubble also correct I'll make doormats from the stubble. Wonderful thought again >> and then I will sell that doormat. And so the idea goes from stubble burning to making doorbats from stubble bought by farmers bought from farmers and from that funding more stubble buying. So, so, so this is what increase in both novelty and clarity of ideas looks like that with a with a gentle non-judgmental AI, you're able to get a student from going from stouble to idea student yet.

17:31 It's not like the AI has come up with this idea. >> Fair enough. >> The AI is only nudging the student to articulate what they already have in them, do it better. So that's that's the journey of the improvement in novelty because risk taking has gone up and has reduced and the quality of articulation that has gone up because and this is happening in Punjabi or this is happening so the language is what you are familiar with >> in terms of in terms of the distinction of scoring and for example of course from that point of view if you could share on that.

18:10 >> So yeah so so there is a the training happens on a golden set right of that hey these are these are highquality articulations these are medium quality articulations these are low quality articulations >> based on parameters like specificity how feasibility etc so so those are factors that are factored in on the training set right and then the then this AI does the evaluation so the AI that is trained is trained on a golden set >> which is different from the AI that's eventually evaluating the ideas that are >> fair enough. Fair.

18:49 >> I wanted to understand this from you. You just mentioned two things which which immediately kind of you know made me want to go a little deeper with this. Everyone talks about critical thinking augment etc and things like that but we should not let go of critical thinking and that's what I completely believe in as well also you mentioned you know articulation yes you're absolutely right because a lot of people nowadays are are right now struggling to let's say find new jobs they might be fabulous in terms of what their skills are But softer skills are missing missing and a lot of organization are looking at that aspect.

19:32 Now when you came up with sarthi it's a tough line to toe sir how did you make sure that you allowed for critical thinking to continue because the way the sense I got it's not the AI giving the idea it is the person interacting with the idea and the augmentation is coming from the chatbot. Am I right? Did I understand that correct? >> Absolutely. It's a very fine line. How did you maintain that? >> So I I think this line or this comes again back from our pedagogical work itself, right? So what we measure our learners on are four Cs. Communication, critical thinking, collaboration, creativity, these are four C's that OECD, WEF, etc. all recognize that keep in an AI abundant world. These are four C's that are needed for our people.

20:25 M and there's a UN so we partnered with MEL University in Canada Montreal they actually help design the metrics to or the rubrics to measure these for us right so there is a school of education in MEL university in Canada they help and they work on the field along with us to come up with communication means these eight sub things critical thinking means these things fair so this >> this underlying understanding of communication, critical thinking, etc.

20:58 is deeply embedded in us both in our content and in our AI. >> Okay? >> Right? That this is what we are trying to build. >> So with that as the goal that the AI understands that you are trying to enable critical thinking in young people, >> right? So you hence cannot do things or you don't do things that will take away that critical thinking. >> Sure. >> So I I think that's at the core of it.

21:22 Then in terms of the how to do it, I think a lot of it is about not giving answers which is reasonably obvious that when a student asks give me an idea, >> the AI does not respond with an idea, right? That the AI asks you questions and engages you in a conversation for you to come up with an answer. So that's like one of the first principles of the AI the way we've trained it. Huh? This is what the AI will do. The AI will never give you answers.

21:50 to AI will actually and this is the exact converse of what you see in in public and popular AI right that a chat GPT or a claude when you ask it a question not only will you give it the answer it'll also tell you the next step can I do this for you >> correct >> right and all you have to do is keep typing yes >> right and it'll h please do h please do h please do and it'll just go on and on and on right that is outsourcing thinking Wonderfully put up.

22:21 >> Here what we're trying to do is that hey we the AI is asking you questions back. So I feel AI is extremely powerful if leveraged in a pedagogically intentional way identify ideas right and then how do you use AI to solve this problem >> where our problem is not that oh can students use AI to come up with ideas that does not help their ideation this is not their final idea in life we want to build the muscle right muscle does not get built if somebody else lifts the weight for you. So I can come and lift the weight.

23:00 >> We have to teach kids to lift the weight of so learning happens through this friction or through this exercise or through this gyming. All right. Absolutely right. So using a machine is fantastic but you use the machine to exercise your body. >> Sure. >> Right. You don't get the machine to do the lifting. >> Correct. >> That's that's really the distinction. fortunate these days it's a skewed it's it's it's getting a skewed in that way which I feel is tragic to be honest mkin and I think you know we need to kind of wake ourselves up to the or maybe it's just the initial excitement and then eventually we will kind of get you know grapple with it enough for us to kind of reach a point >> as a both as a technologist and an educator I do worry because unlike most other technologies u this technology is extremely widespread and easily accessible right I have 17 and a 15 year old kids right and it's very easy for them to outsource thinking out outsource imagination so for them to not build this muscles for them to not do this is very easy yeah >> and I don't think today there is enough conversation you're Right.

24:26 Right. Correct. You are not designing education for doing things that kids like. And that's one key example of like hey agency coming up right that hey when students are doing things that they enjoy that they want to do they are engaged right but if they are having to do homework that they don't want to do >> they will I agree >> we used to do it to be honest me if you remember if you didn't like a subject we would figure out ways and means to go ahead and just kind so I mean yeah we've all done it So the instinct remains the same. You're just going ahead and well I should not be saying it but kind of a situation but absolutely you know and it's scary was the friend who would be right from you.

25:29 >> So and again in copying it no learning was happening. >> Yeah you're absolutely right. So now that coping has become extremely easy. >> But tell me this then I mean these two things come to mind. I will come to the the fact that and I think you're doing a wonderful job in terms of working with governments because I think that probably is a way forward. I'll come to that but prior to that I want to understand probably the the safeguards so to say like we just were talking etc things like that in terms of the AI as far as arti is concerned how do you know or or is there a safeguard in a way that maybe a student cannot copy from what an AI is and just paste it for example that is just one thought that comes to my mind the other one is what is sarti is is programmed to refuse in terms of the perspective and a conversation because I feel that's also very important.

26:24 >> Yeah. Yeah. Yeah. Huh. I I think both are important questions. right. So on the I'll start with the second one maybe go the first one that >> so first is everything that is criminal that is unlawful. So if you want to if a student asks it to do say to sell drugs right etc. So everything that's unlawful etc shati will refuse everything that shati and we've used we've used libraries around what is child safe so what's not seen as child safe like pornography or fair right >> those are all things that >> sati would refuse so combination of criminal and not safe for kids are the two things that sati would point blank refuse conversation engage I think on the first one is there >> right. So like will Sati refuse it? No, Sati will not refuse it. Sati is like Sati will try to nudge you towards how do you make it yours right? So for example if you submit if you submit the idea that oh I want to my animal u like one of the common ideas that we see is around protecting farms from animals. Okay.

27:45 >> We work in we work in rural parts of the country, Punjab, in Hana, in Uttakra. It's a large problem. >> Students come up with >> Kerala is like rampant with that issue. >> Yeah. So, so it's a large problem. So, students come up with ideas around this and it's hard to know if this idea has been copied, if this idea was created by Chad GPT or created by another friend. >> The point then is that the the AI asks you, okay, what are the farm sizes you're looking at? Right.

28:14 Is it given you given AI knows that you are in Utra is it a step is it and hence how will some of these works so the intention is to get into a conversation it is not like homework it's like right let's make this your idea with more inputs from the learner and I I think that's a key question that if a student is not motivated right and if they just want to if they just want to copy paste and get away. You will get away. At least the education system still doesn't have ways to allow you to like not get away.

28:52 >> Our point is not to penalize or punish those students, but to actually excite and engage them. That's fine. You're starting from here. >> How do you make it your own? Right? How do you make it something that is relevant? And what about you? Our our hope like one of the experiments this year is to build understand student personas that if a student is a lot more interested in sports >> right >> then can you have like oh when you're talking about a farm and farm protection to then farm looks like a boundary or so you use cricket terminology correct in so which can then excite the student so how do you bring bring motivation and excitement and personalization into this rather Then how do you prevent and protect and dissolve? That's broadly the difference in approaches.

29:43 >> You work a lot with schools I'm pretty sure. Meck and as you just mentioned you know the UP story which you talked about which unfortunately was quite sad but tell me this before of course I move on to the state government Andhra Pradesh is one where you have an agreement etc. and I will come to that but personally speaking as far as mech is concerned what have been your findings in terms of the the let's say the downside of of AI in terms of the way it is being you know patronized by students or probably even schools in a big way. I remember once I was having a conversation with someone and and there was this entire thing about the fact that where teachers are using AI to come up with question papers and etc and then the students are answering through AI and then the teachers are checking through AI. So I'm like okay so one AI is having a conversation with the other AI and everyone's else seems to be comfortably numb in the in the middle.

30:44 So I just don't want to talk about what the issues are. Of course in terms of you will highlight them but according to you what could be the probable solutions as well. >> Yeah I so I I feel like you've touched upon the issues fairly well. I'll just broaden them that most of our education has been designed around assessing for knowledge. >> Right. >> Right. That do you know something? and >> all that assessment happens through language. Right. And these are large language models which then are actually very good at language and very and have become very good at knowledge.

31:19 >> So there is no way in which it's going to be possible for us to compete on knowledge right with AI. >> Yeah. Yeah. >> Agreed. >> But there are enough and more human endeavors that are beyond just language and knowing something right. So courage to do something or actual doing is very different from just knowing. Right? You might know that oh this is the way I can paint right but to actually >> requires you to pick up the brush and do it and today and maybe it'll take robotics sometimes before that also gets completely automated etc. But there is a long way to go before hey a teacher coming into class and inspiring students will become obsolete. I feel like hey that is extremely important.

32:06 >> That would be a very dark day unfortunately if that happens. I know but I I don't see it happening. I don't see it happening in the short term for sure. Right. This idea that oh AI will do everything I feel is is I think there is we have to remember that it's in the bits and bites and language and knowledge space that AI excels at. >> But there is a much much bigger physical world. I feel the sooner we move from what a person knows to what a person can do right like how do you leverage and use the knowledge and the tools to be able to do stuff to be able to build stuff to be able to solve problems to be able to create new things. I think so our education system our assessments have not moved to that we continue to stay with do you know the capital of right when was this war fought >> more like road learning in a way. Yeah, you're right. Google right now Google you couldn't copy paste a long question. Now the chat GPT will do a very good job of a long question it will give you the answer.

33:14 >> but we haven't reimagined and most education systems haven't reimagined that what can human plus AI together do. >> So if we start thinking of AI as the calculator that you had in your engineering math exam, >> right? Hey, I should be able to use AI to do these computations for me. I have to use my head to know when to use the calculator and when not to use the calculator. Like when should I come up with my fresh ideas >> and when should I possibly say hey this idea problems maybe AI can do a good job of analyzing that hey maybe this is a way to think about it and then that pushes you to improve it. So this combination of human plus AI and education being designed for assuming that AI is available.

34:00 >> Yeah. >> How can human beings actually use AI to create to solve to do more? I feel like the sooner education shifts to that. So to your question of how do we solve for it? That's my opinion on how do we solve >> then logically that that brings me to this that leads to this question. you know when when when I was reading up on udyam Andhra Pradesh government me I want to talk to you about that from the point of view of the fact that you know it's a it's a it's a state government formal agreement that you have which is wonderful by the way now I want to then logically ask you I'm sure you want to impress upon the government this very novel thought that there needs to be a reimagining in terms of the entire policy and probably the system what did a state for example still ask you to change that you did not want to.

34:51 One is that and second what happens to Sari when for example a state government changes critical points both according to me >> tough tough questions for both and I so back to the original question which is hey how do you want state government so we just talked about a visualization of this is what education should become >> but education is extremely complicated and complex and old and there are set structures so the way udam has gone about introducing this change into state governments is by helping the state government create a new subject.

35:27 >> Okay. >> So with every state so in Uttarakhand this subject is called Koshalam. Hana calls it Kushal business challenge. Punjab calls it entrepreneurship mindset curriculum. >> Andhra Pradesh I'm trying to remember what's the name of the Andhra Pradesh curriculum. I think it's it's very similar to the Koshalam name in in Andhra Pradesh. So these are curriculums that the state has created with inputs and work with us which are taught two days or two periods in a week. Nice that because you >> and these are mandatory these are not selective or anything mandatory.

36:03 >> These are so every student in Punjab, Hiana, Uttakand grades 9 to 12 and in Pradesh grades 8 and 9 goes through this. If you are in a government school, this should be happening. Like with all government school systems or government schemes, execution may not be 100%. In Uttarakhand, we've been able to make execution go all the way up to 75 80%. In some other states, we are working to improve that, right? But by design, this is supposed to be happening in every school, grades 9 to 12, two periods a week in these Punjab, Utraana, those are the most mature states for us, >> right? So that was the what Punjab has done additionally right and this is to your question that what happens when the government changes because we worked with Delhi and then in Delhi government changed.

36:50 >> Yeah. >> Right. and there was this entrepreneurship mindset curriculum that was created in 2019. It lasted all the way till 2024 till the government lasted. After that the new government has kept entrepreneurship but has morphed it slightly from being something that was available to everybody to something that's for the few that hey some bright ideas get 20,000 rupees in the new scheme which is also a good way but we had thought about that hey can this be available for every child again going back to potential how do you enable this >> what Punjab has done which is the same leadership that was in Delhi is that it's made entrepreneurship a mandatory subject for 11th and 12th.

37:33 >> It's now part of your board exam report card. So not only >> so not only the government schools but also private schools that are affiliated to the Punjab board are now teaching this. >> So we as UTIM are helping affordable private schools in Punjab. This is how you can teach the entrepreneurship subject, >> right? That that students have to appear for, >> right? And of course, we do this pro bono because for us, we are a nonprofit organization and this is not something that we want to make money out of, but we want young people to be able to acquire and equip themselves.

38:09 >> But I broad question key one, make it good enough that the next government also wants to keep it. >> So some of the things that we've done include >> Yeah. Because I mean there's no vested interest to be honest. That's the first thought that comes to my mind. I I'll give you I'll give you counters right. in one of the states and let me not name the state there was a new is officer that had adopted that had taken charge of education and I went and met her and she said mckin in my state my 10th pass rate is 15% lower than the national pass rate. M and I'm like that's a fair challenge. I I did tell her that madam your job does not end at 10th pass rate which is true >> as a believer of focus I get it upth right your state students deserve right to get there and you should work on ensuring that that happens but please realize that these kids even after they have passed 10th right there aren't jobs they will need to equip themselves So, so but now to her and to many many others like passing 10th grade or passing 12th grade or trying to get into engineering which are the norms for most kids parents, students, teachers culturally start earning a good salary. the reality that that dream is available only to 2 to 5% right we are unwilling to accept >> accept reality and we are okay and I feel like this is a cultural problem that we are okay with 2% doing well >> we are not imagining an India >> where 90% does well >> right hence we have this uninclusive prosperity there are a few I have been very very lucky flipkart succeeded so I am amongst the very few who have done really well >> but many of my batchmates who like me did not get into YIT who felt like a failure at 17 have not done well right and the system disregards them and it's like you were not good enough which I feel is not true I feel I agree but cultural shift it is a mindset perspective or Megan let's be honest with initiative like yourselves with like like you have gone ahead and initiated and the fact that you're constantly very passionately involved and you're constantly chipping away it will make a difference but it will happen according to me yeah absolutely sir and I'm praying for it and I'm and I'm giving all the positive energy as I can for for this to yeah in a big way for for it to kind of go ahead and truly come out with flying colors but tell me this in terms The parents you mentioned parents acceptance 2 to 2.5% you know that thought that when you said that you're lucky and I'm very lucky etc buffet line was it called the utarian lottery or was it lottery ovarian lottery yes so we should beh correct >> I wrote in mudam's birth note that the ovarian lottery is unfair I don't know why it is probably linked to that entire perspective but parents I want to understand Mckin how are they accepting it because see let's face it any society across the world looks at results you know like you said 75 to 80% utrahan for example how many ideas are getting converted or sponsored for that matter or are kind of finding a sense of life once it's kind have done and people are genuinely kind of have something palpable and physical to tangible to kind of go ahead and say okay this was my thought and today it's a business for example >> yeah yeah no again tough question and the reality is not is not great that what it takes to go from a business project in 11th and 12th grade to be an entrepreneur it's a long journey >> we right now have on an average about 300 odd entrepreneurs or ideas emerging from places like Delhi, Punjab, Uttra etc. per year which is a fairly small number but as role models >> these 300 are catalytic right I'll tell you one quick story of this entrepreneur from Delhi guy called Nick Hill >> so 11th May when this project was created when he had to go through this project called business blasters his idea was to create a stick for the blind right >> he was very happy >> yeah he was very unhappy with the controls under the thumb of the blind etc. But he could not find good controls.

43:31 After a few months, somebody introduced him to 3D printing. 3D printing technology. He he made it and he loved 3D printing. >> He fell in love with the technology itself. He changed his business idea from stick for the blind to building 3D printers. >> Oh. He then went and built a 3D printer all by himself. This is he was in 12th grade then 12th grade school righted made new additions eventually made nine 3D printers not a single one sold. And then it dawned on him right and this is what life and entrepreneurship teaches you that it so and the moment he pivoted to this idea he had enough and more business right this kid has sold to Israel >> has sold to ICMR wonderful >> so he has he has two and a half to three lakh rupees a month as an income Yeah. from just selling 3D printing as a service. She has 30 3D printers right now and this is a government school kid. So I I think the reason I brought this story up is K. If you allow young >> I thank you that you brought up this story. It's a wonderful story. Very inspirational.

45:00 >> Huh? >> Was the name of the company by the way? >> it's called Nova Nova 3KD. I'll I'll send it to you. >> Yeah, please do. and and and for everyone who's watching, let's go ahead and actually appreciate and and patronize this company in a big way. It'll be so fabulous. You know why I'm saying this also is because I remember I was having a conversation with Dr. Kamakoti and he said something which I thought was so true. He says yeah startup people should give them their first purchase order. It'll only inspire them.

45:33 It'll only motivate them. >> Yeah. So, Novanic 3D is 3D. >> No. Okay. Wonderful. He's at four lakh average monthly revenue right now. >> Great. >> 35 3D printers. >> Great. >> Fabulous. Right. You were asking me, right? The number is small, right? And it takes it takes a lot to be an entrepreneur. That is reality. But there are people like Nikil who are emerging.

46:05 There are people like Ashish who's built a a logistics company and now has like had three crores of revenue last year. >> Wow. Wonderful. >> Doing logistics between Bihar and Delhi. Very very simple they business or second year. So, so this I feel is possible that hey and then when these kids go back to their school right teachers principal correct like are so deeply inspired right >> that's what I was trying to get at >> so so that is I think that wheel is slowly starting to turn the parents we see both with parents and with teachers one of the things that shifted was when I think three years back we started pushing for every school to do a school mea >> that at the end of the year let the students put up stalls of their creations >> and the number of parents and teachers who have gone on record and told us wonderful because we think of these kids that academically weak because the schools 's definition or the parents definition of productive >> is only related to copy, >> right? And now that you've created one more avenue for them, these kids are able to solve problems for the community like do things that are like correct.

47:40 So I I feel that's that has shifted respect or acceptance because you're right. Most parents initially exactly which is also why we like hey this is not all of education. Education continues. So it's a little bit that at least exposes you towards a life like this. >> It gives you the opportunity to go ahead and fly. And as far as you know achieving your dreams is concerned. Tell me this if Meck I mean this came to my mind immediately when you were going ahead and talking about you know the wheel has started to turn. So for example if you want to actually go ahead and make the wheel turn faster maybe I will hawk back to your collaboration for example with Miguel University as you just mentioned mentioned earlier. So is there anything that you're probably doing from the point of view of maybe some VCs that you want to probably bring on board as far as udy is concerned or sarti is concerned etc which might just probably assist school for example how doing a mea for a lot of the VCs maybe that might help I don't know just thinking aloud with you >> no thank you we've we've tried some of this that we've done state mailas that the best projects in a state so punjab hala Uttan all of these states bring up the top and it becomes competitive >> it's like sports and then for the state mil VCs and entrepreneurs have shown up and back some of these students >> I think along with money what is really needed is handholding and advice >> exactly just today morning I created a WhatsApp group where Suraj who is one of the co-founders of delivery right he agreed to mentor Ashish wonder the guy that I just spoke about business Suraj is like yes to connect I'll talk >> right >> and I feel like the entrepreneurship community in India is more than willing to help these young kids because when you meet these young kids right the energy you feel the I'm sure >> the potential and possibility you feel they'll go far can we please come together and support them >> yeah so that is on already something that you are constantly working with that's very refreshing to know >> we at the same time we don't want to take entrepreneurship away or the difficulty of entrepreneurship away >> yeah I absolutely agree with you >> right >> sir because I'm a very firm believer in that >> which is true I want to now very quickly understand from the you know the pricing point of view for example openai at the end of the day Meckin gave give you credits and technical help through the accelerated. Now one very quickly which model runs Sarthi and what happens if its price or terms change because that's a big one no in terms of these days especially with tokenization etc all those things being talked about constantly what does it cost per student for example a year >> yeah so we are we are currently at a level where our total AI cost so there are three underlying models used there is quite a bit of Gemini used so Gemini is the primary model that we end up using And there is open air used our monthly cost of them is about two lakh rupees a month.

51:04 >> Well, that's not bad. >> Exactly. So, we're nowhere at a point where this price is worrying us. Okay. >> So, at about two rupees a month reaching and serving seven lakh kids, it is about 4 to 5 rupees per child. >> Our overall cost of the program is about 100 rupees per child. >> Right? So the technology cost is still about 5% of the overall program cost and the value that the technology brings the results that you already saw is already quite compelling. So we not worried and I'm hopeful that both from how pricing wars are going on prices will only come down that better models will lead to prices coming down and AI doing more >> and I'm also hopeful that most of the AI organizations are now increasingly wanting to do philanthropy are increasingly wanting to solve key are how can we help AI have productive and useful things come out.

52:09 >> Yeah, fair point. Okay. And now as far as the rules are concerned, I need to obviously request you to probably explain that Meckin in terms of let's say the DPDP act. Now children's data gets special treatment under India's new data rules. So >> whose responsibility is a student's WhatsApp chat? Is it yours? Is it the schools or for that matter the states? >> Yeah. So currently it is currently that responsibility lies with us. Okay. we are working to help transition this over to the state >> right that these are state government kids and some of the state governments are actually more than willing in some cases there isn't yet both understanding and capacity and willingness on their end. so we are doing a little bit of a build operate transfer that's the eventual goal >> right that hey let us do the innovation build this out while keeping the state government completely in loop and them being part of our agreements okay >> and then transfer this ownership back to the state because the curriculum for example is already part of state the state agreed right >> the only reason right now this is this is running on our servers or we are managing and maintaining this is because their current there isn't current capacity and I see this a 3 to 4 year journey where we are able to then transition this over to the state.

53:31 >> Okay. Very quickly I I remember when I when I was trying to read up as far as Sari was concerned etc. this came to mind which I wanted to then ask you. You say this mechan that it's a case study actually your case study says this that student data is masked before it reaches the model. So essentially then I wanted to ask you what does the model still see? >> The model the model sees attributes about the student that the student is a male is coming from Uttarakhand is in this grade right. So descriptors about the student but PII personally identifiable information is not.

54:11 >> So the name of the student the like specifics like Aadhaar or mobile number is not shared right. So this the model has no way of contacting or communicating or >> like personalizing based on that like >> fair point. Okay. >> WhatsApp is essentially the engine right M which is essentially >> WhatsApp is the entry point. WhatsApp is the distribution channel. So I mean this is essentially as we all know something like that maybe happens god forbid and and meta has to change policy for example or price >> I think WhatsApp currently I don't think we have thought of a backup right >> how about something like from by zoho Absolutely. So I I would hope that we are able to migrate to that. The issue Sudhi is that this as in it needs mass adoption.

55:12 >> The reason we use WhatsApp is because kids use WhatsApp. >> Correct. >> And for kids to use an attack, right? So if they start using it, we are very happy to. And I think that is where you know the Yeah, I agree with you there. >> The the chicken came first or the egg because that is where the problem is with Aratai also. If I want to use it, maybe you're not using it. I can't connect with you >> then. Exactly. Right. I think till that gets solved, we are a small nonprofit and we are not if I were building Flipkart, sure I would I would build adoption.

55:43 >> Correct. >> Right. I would get you because I have multi-billion dollar funding to be able to influence user behavior. >> Here a small whatever $4 million annual burn nonprofit, right? Trying to do all of this at scale. We do not have marketing muscle to drive adoption of a new tech product, right? Oh, we'd rather use what users use. Yeah, users use WhatsApp. >> No, this this came to my mind. That's why I thought I'll ask you and maybe it's food for thought. Maybe something that you'll probably be already working on or probably will be moving forward.

56:16 >> I think technical architecture wise we are ready, right? WhatsApp. There is some other messaging platform. Sure, we can move it. As long as it offers the same tech specs and tech capabilities, we we'll be a we'll not find it hard to move. Fair point. Very quickly, I need to of course now understand this from you. In conclusion, I could really truly carry on with this conversation. It's been fabulous to be honest. Me, one thing before I I I I conclude this and I would request you next time around.

56:44 It'll be great if you and maybe one of a a great success story. Yes. Can both be a part of Front Page. It'll be so so amazing. We'll be we'll be more than happy >> and and graceful about hosting the both of you and that's again something which I'd love to do. But >> very quickly I'd like to understand this. You mentioned marketing right now for example Noonic 3D for example or the company which is the delivery company logistics company for example. Why are these stories not more popular sir mckin?

57:19 >> I mean everyone should know. I mean, I'm sure a lot of people once they know it will gravitate towards you guys to assist in a big way. >> Yeah. Yeah. No, thank you. And I I think so. Some of them have gotten covered. There are YouTube videos, there are news stories about some of them. but yeah, I I do think that hey, capturing media attention or capturing like what it takes is hard. at least we don't know that art, right? you are we are happy to be guided and happy to be mentored around how can some of these stories reach where they deserve. Yeah.

57:53 >> so I I'd love to get get suggestions and ideas on what's possible. We have we easily have 30 40 very very inspiring >> right who which like a lot of people would love to listen to and understand and just feel happy about this is happening dured So very happy to like create a series, help work with you or if there are other people who you think we should reach out to. We would love to create more visibility to these students. I am busy right now helping them solve problems like the delivery marketing.

58:38 Don't worry, we'll try our best. I'll start right now for everyone who's watching. as far as Mechin is concerned please reach out in the comments below. We at Front Page will be more than happy to share the details as far as Mechin and his organization are concerned. You can find them on the net. But please go out there and actually make a difference. If you can mentor those kids, it'll be fabulous. If you can fund them, it'll be wonderful.

59:03 If you can just go ahead and be a part of something that Mecc is going ahead and struggling with, that will help in a big way. So Megan, this is just my very humble initiation right now. Let me think of more ideas as we keep moving along. >> And if you like those stories, share those stories, right? Fabulous. Yes. >> Just those ideas spreading that oh young kids can do this. >> I feel like young India has immense potential and immense entrepreneurial potential. So helping more and more people see and believe young people can do this, right? Enable young people around them is the change I feel our country needs. truly inspirational.

59:39 Finally, one message that you'd like to give Meckin as far as the young entrepreneurs out there are concerned. What's that one thing that you'd like to give which is probably something that you learned the hard way? >> No, don't let anyone tell you that you can't do this, >> right? >> believe in yourself. Go after it. don't worry so much about acceptance right it'll come just go and do as in believe in what you want believe that this is doable and just go and do >> wonderfully put sir mckin I truly wish you and your entire team all of the best and the fact that that invitation will remain open for you to join us again and it'll be truly an honor to have you back with one of your glorious shining stars and anyone else of course also who you want to go ahead and bring on to the show. It'll be fabulous and wishing you all of the positive energy moving forward. You're doing some fabulous work. Please continue to do so.

60:45 >> Thank you so much Si. I will take you up on the invitation soon. Thank you. >> Absolutely. Well, there you have it ladies and gentlemen. Honestly, what an enthralling conversation and a conversation which is so important, so timely and needs to be had again and again and again. And I would really hope and pray that you will please give us your feedback and actually probably reach out to Mecc on this in terms of the true impact that he's trying to create. You could be that one helping hand or maybe more to assist in this wonderful agency and enterprise moving forward. This is front page by the IM network. Like, share, subscribe and always remember think AI. Think I am