Transcript 0:00 Education is the one thing that we can give back to society. I was afraid as well. Today, we have over 50 million students using it. Now we can start thinking about changing the system. The US wants to keep its advance. 0:16 They're gonna have- [on hold music] 0:25 Today, I'm very honored to talk to Niro Khosla, and she is an entrepreneur and philanthropist known for her work in education and, um, social impact. 0:35 She's the co-founder of CK-12 Foundation, which provides free customizable learning materials to students and teachers worldwide. 0:43 And I'm very passionate about education, especially in this new era of AI, so thank you so much for joining me. Well, thank you for having me on this, um, you know, this conversation. 0:54 It's a podcast that's called Inference, so inference in the sense that we're making conclusions about what's happening, so- Well, thank you for inviting me to Inference. I mean, it's, uh... 1:05 I love to talk to people about what I'm passionate about. For me, education is the one thing that we can give back to society, no matter what. 1:16 We can create all the, you know, ways that we can make financial freedom- Mm-hmm... but that can only happen if we actually truly help our young students to know what they need to know and how to use that. 1:34 How does the last three years that has been pretty disruptive for many- Yes... industries, how did it change your perspective on education, your perspective on CK Found- the CK-12 Foundation? 1:47 So when we started, we were trying to figure out how might we make education available and meaningful for all children 'cause remember, you know, what you got in a textbook was so one-sided that you got what you got. 2:02 Mm-hmm. So I wanna be... I know that true learning can happen only if you can, you know, create your own connections to these things. Mm-hmm. 2:13 As technology was coming along in the 2000, early 2000s, I knew that this was gonna go far, and so I said this would be one way we can actually get to larger number of students and help them out. 2:27 From the very get-go, we were free. So technology has brought a lot of that capability, the multimodality that's needed in learning. Specifically in a textbook, you cannot just read or a video- Right... 2:42 just watch, right? You cannot make sense of that unless that concept you're learning about is in the best modality that you can learn from. So imagine physics, elec-electricity. 2:57 I learned by reading a textbook how, you know, electrons move and, you know, kind of- And you just see it. Yeah. So now we have simulations. We have ways for them to really kind of get... 3:10 And now, you know, a lot of these things are in every home. So we wanted to make sure that we get every kind of modality, and that's what technology offered us. 3:20 But we did not know what was going on in the minds of the students. We did not know that they really got what we were trying to show them or teach them. 3:32 So it was only when AI came along in the last three years that we realized that there was an opportunity. Hmm. It wasn't perfect. It's like every child, every technology in history has never come ready 100% to use. 3:49 Mm-hmm. The only way you can make technology mature is by iterate, uh, uh, iterating it, right? And that requires all of us to join. 4:01 So if you think about what happened in the last three years, there's been, you know, like, OpenAI got million users within a few days or whatever the timeframe was. Yeah. Short as heck, right? 4:15 So when you think about that, something big was happening, and a lot of people were afraid, "Oh, my God, we're gonna become machines," or, "Oh, God, the teach- students are gonna cheat." 4:26 So for me, it was these three years, admittedly, I was afraid as well. Interesting. Okay. I could see things, but I also noticed that, you know, prompting by itself is a great learning tool. True. Absolutely. Okay. 4:43 It's the, the, the skill of asking questions. Yes. 4:46 So one of the things that I kept telling people, you know, don't be afraid because if your students can ask or prompt the machine to get to where they want it to go, they are thinking about this. 5:02 They're thinking about concepts. They're thinking about learning. They're thinking about how it applies. Am I right? Am I wrong? 5:08 They may not know that, but they know when it answers, "Well, that doesn't sound right," or, "I wanna get deeper," or whatever. So you keep prompting. If we can teach these ki- young kids to question- 100%... 5:22 the whys, right? The five whys that been, for a long time we've been following. I think that itself is going to be a huge... It was going to be a huge... 5:31 So that last three years, we introduced, on top of our learning sciences, learning development sciences, all the concept, uh, uh, um, maps we had done, all the connections we had done. 5:46 So all the thing we had done at the back end of our system, we saw the next step- Mm-hmm... which was the chat. Two and a half years ago, we introduced Flexi- Mm-hmm... our, uh, student tutor. 6:00 Today, we have over 50 million students using it. Oh, wow. And how are the results? We have over 100 And 50 million questions asked. 6:14 The moment we started doing this, I told the team, "Guys, we've got to categorize, tag these questions they're asking." Okay, some, many of them were just procedural questions. "Help me," you know, "solve this. 6:27 Help me understand this, define this," you know, kind of questions. And th- those are fine, right? But the ones that really got me was a ques- questions like, "How does the sun burn when there's no oxygen in space?" 6:43 That's coming from a young kid. You've got to kind of say, "What's going on?" 6:49 It's definitely, you can deduce from that question this child or this learner doesn't realize that it's not a combustion- Mm-hmm... that's happening it's... in the sun. It's actually a nuclear fusion. 7:06 You can correct that. It's these kinds of situ- you know, kind of w- you know, kind of questions that lead to, you know, us deeply understanding students and really helping them. And that's what we built, 7:21 a platform that connects the concepts, the g- content. We built flexbooks. The flexbooks were nothing but flexible, you know, content- Mm... that could be personalized to the level of each and the language of each. 7:38 And the, you know, just standards or whatever else you wanted to do, you could do. Your classroom, the teacher. That was the three years. Uh, we were already doing this. 7:50 The thing is just knowing that the question is right or wrong doesn't tell anybody anything. So standardized testing, you base it on questions right or wrong. It doesn't develop- You're not getting it... your thinking. 8:02 Right? But now we can deep knowledge trace this where it went wrong. Mm-hmm. How it went wrong. That's very, very interesting. I have five children, and they are in a... You have four? I have four, yes. 8:15 Yeah, that's amazing. And what is the youngest age? 33. Oh. [laughs] So it's already, it's already a little bit late. But mine are still in public school. Yes, yes. 8:26 I just see how early we still are with AI because- Yes... public schools just cannot catch up that fast. Yes. And my question is, do we need to rethink the whole education system from ground up? 8:41 How do you think about it? You know, when I started, I used to say, "Let's just do this without..." I didn't go for adoption in any of the public schools. 8:50 First of all, they didn't want to adopt because we were free, right? We're still free 18 years lat- uh, l- later, and we're still going to be free for as, as long as people want us. We'll find ways. 9:01 What was surprising is that for adoption, they had money. Schools had money. Districts had money. States had money, right? They didn't want to lose that money. Yeah. 9:13 So that's the legacy system- That's a politics, yeah... we were dealing with. So I wanted to do a kind of a Trojan horse approach where we could level it to each kid, kind of figure out from... We built in the very... 9:29 Within couple of years we started, we built, started building the heart of the system, which is our adaptive system. And that system we could tell where the students are missing things. 9:42 And so for us, that was all about starting from very early doing, starting doing that. Mm-hmm. 9:48 So in essence, what the teachers, schools, districts have to do is to show they're succeeding, and the measure of success really is just standardized testing. 9:59 And that's a very, very deep-held notion in every teacher, school, et cetera today. Then, you know, we started doing it for the teachers and kind of helping them, uh, help every child. 10:12 But soon I said, "You know, maybe we should just work on the students so that we can disrupt that system, right, in ways..." But it's very unnerving. 10:22 So you have to find ways where as you're walking, you lift up your foot where the f- you walk first, uh, first foot out, lift up your second foot, and then you decide where you're gonna place it so it's stable and continue that walk, right? 10:38 So that you don't- Mm-hmm. So changing it completely is a pretty hard thing to do. 10:44 But I think as AI, the last three years have given us tools that now we can start thinking about changing the system, introducing, hey, which we already do. 10:56 We have teacher assistant, which actually tells the teacher where every student is, how are they gonna progress in the next thing you're gonna teach, and so that, you know, they're aware. 11:09 And that's gonna be a shift to the change. Do you follow teachers', uh, questions, what they might need from AI? What's their interest in AI? Oh, we've been talking to a lot of teachers. We've been working with teachers. 11:21 We've been working. And so, you know, yes, and we get lots of requests. But you know what it is? It's very hard to imagine. There are very many teachers who can do that, who foresee things, who can see. 11:37 But there are others who are very afraid, tired, and kind of not ready to take risks. So you have to deal with all the different kinds of profiles, and that's what we're trying to do. 11:51 That's why we've been around 20 years. I have, you know, almost 20 years and, you know, I've been thinking about how do we change the system. We have to. It's a long-term view. Hopefully not... 12:03 I'd say another five more years when things really start changing. On the systemic- Yeah... level. Yeah. Yeah. There is still a lot of fear. You know, I like our work with India today. 12:14 We're trying to work with India to see if we can really change, and they're very pro this change. 12:21 So if we can get people to see the change and it's not scary, yes, there'll always be bad players, but we have to take care of these bad players. 12:30 Like when we created content, we had domain experts, we had human beings keep checking everything. What are the s- systems saying to the student? 12:40 What are the stu- You know, keep doing that, the quality assurance, quality, um, you know, control. And now, as automation comes in, you can control because with attention you need, you could teach machines. 12:56 What skills do you think are essential in this new AI world for teachers and for kids? Teachers have to have an open mind. They have to think about what's possible. 13:09 When the internet came along, you know, when your, um, um, what do you call it? Calculator came along. Oh, they're gonna cheat. They won't know the how to do the math. Yeah, sure, but you know, 13:20 some of these was not necessary. I, I don't say you don't need to know how to compute or how to do math, you do. But there are numbers so big that you can't do it in your brain. 13:31 So similarly, when, you know, the internet came along, "Oh, they're gonna cheat. Oh, this is dangerous," blah, blah. Same story continues. 13:39 Fifth Avenue in New York in the nine- uh, 1800s, late 1800s, they had, uh, or ninth, early... I can't remember, 1800s. They had horse and carriage, right? 13:53 And, and rainy days, they would stuck with mounds of dung and, uh, horse manure, right? It was... They didn't wanna do cars. But they came along and people start saying when you show, keep at it. 14:07 Anyhow, so I think teachers will, we will have to educate them, which we are doing, and show them how the system works. Right now, they don't really understand how the, this works. 14:19 I think they're all smart enough, they will get it. With students, students need to understand they're going to have to learn many things. They have to be generalists one from day one. 14:31 They're gonna have to learn how to write, how to, you know, do math, numeracy, li- literacy, all those things you have to start learning. We never thought English majors were going to be much of just... 14:43 But look at it today. Prompting comes best- Yeah... for people who understand language. That's true. This is where I think we are getting there. 14:51 We have to start, you know, kind of also pointing out to the students, like in our system, you say, "What do you wanna become?" And if they say, "I wanna become a stage, uh, designer for plays and stuff for- I know... 15:04 right? And I don't need math and science. But wait, how are you gonna know about lights? 15:10 How are you going to know about areas, what to put where, how does these, what kind of shapes you need to put so that you can do best efficient use of the stage? You do not need math and science. 15:22 You need creativity, you need critical thinking, you need collaboration, you need communication, and you know, you need community. 15:31 And so that's some of the things that we've, you know, haven't almost got all of them down pat, but they are starting to appear. And I think many people who are offering it in a serious way are thinking about that. 15:46 What do you think the companies who offer AI education currently, what do they misunderstand about kids and teachers? I think one of the biggest misconception about AI is that I'll just ask it a question and it'll do. 16:03 Reality is what attention is all you need did, a paper from Google did, was it looked for the best fit probabilistically. 16:16 And now we have to understand it's not that we just do whatever LLMs are doing. We have to keep ahead of them because there are non-negotiables in learning. 16:27 And to be able to do that, those non-negotiables, computers or, or, I'm so sorry, the LLMs and AI isn't ready for a lot of that. 16:37 I think we have to treat every vertical and, you know, kind of continue in that vertical the things that are very unique to those verticals. And what verticals do you mean? By verticals, I mean domains like education. 16:52 Like these are different verticals. They go on top of these LLMs or Gen AIs because they're not AGI as they call it and what does AGI mean? We don't even know that yet. 17:04 Can they really replicate a human being en- entirety? 17:09 Yeah, I think in the, in, in the, in this new era, AI is becoming as important as reading and writing that becomes new literacy point, but we're kind of not there yet. We, we keep getting there- Yes. 17:24 And you, your is one of the projects that get- getting us closer. 17:27 And the other thing I was thinking that maybe sometimes we should maybe follow kids and how they learn technology, just letting their imagination run free. Because they immediately, you know, get the grasp of it. 17:41 I saw that when my kids were in, uh, middle school. At that point, one of the kid- things that kids do, they also have a social graph of sorts. 17:52 And they've always had that whether it was, you know, whatever media we had- Mm-hmm... kids had their own social graph, which means, you know, who connects to who and how they learn, can they teach me? 18:05 You know, that's how things in the children's world. And I spent about fifteen plus years as the head of the education. By the way, just so you know, I was a molecular biologist. Yeah. I, I, I read that. 18:20 [chuckles] How did you pivot to the education? Well, I, I was doing, uh, research at Stanford on oncogenes, and from those, I realized, became preg-pregnant with my first kid. 18:33 I said, "Oh my God- Can't be around radioactivity. So we moved on. I had four kids. After that, I said, "What do I do with my kids? Education." 18:42 So that's when I found the best school that I thought had the best education, which was child-centric. Sent my children there, and we kind of got involved. It was so fascinating. 18:55 I spent about 10 years in the classrooms- Oh, wow. Every day looking at what teachers were doing, what students were doing. 19:04 I would sit down, uh, you know, it was mostly child-centered, so it's child sitting in the corner by himself. Mm-hmm. I would go sit with him and say, "Hey, Ari, what's going on?" And he was shy, and he was kind of... 19:17 And he would keep asking, saying... All he needed was somebody to tell him, "You're doing fine. You're doing it," and help him if he was stuck somewhere, not to do it for him. 19:27 And that's how I became interested in education. And I said, "This is fantastic education. Why can't we give this for every child?" That was the genesis of CK-12. Oh, what does CK-12 C stand for? You should ask me. 19:42 Yeah, what CK stands for? [laughs] Well, K-12 is K-12. C stands for every child. Creativity, critical thinking, collaboration, you know, all the C's- Mm-hmm... that you need or you need. 19:55 Those are non-negotiables for learning, deep learning. How interesting. You thought, I mean, they are very important skills, but they become even more important now, those exact skills. That's right. 20:07 It, it's interesting that you said in the beginning that because it's free, schools could not really get it because of the, their politics. Well, not just politics, also the impression free was junk. Right. Yeah. 20:22 That's, that's also true. Phil- so philanthropy is, is, is so important. But how to actually make it work in education when it's so hard to go with a free product to get inside with that? 20:35 Well, it's, it's, it's, again, most free products don't last. We've lasted about 20 years almost. Right. 'Cause part of it is all... And actually, that's true of for-profit companies too. 20:47 They can continue as long as they can have funds to continue. Funds are very important. That's why my husband, Vinod, and I have stood behind this. We fund it so that we're funding every child. 21:01 And is it like a, how do children get into the program? They can just join it. It's just open. For everyone? Yeah. You sign on. 21:11 And if you're under 13, you would need a teacher to sign you on or your parent's permission kind of. Uh, that's one thing that breaks my heart. You're under 13, you cannot learn on CK-12. I mean, can you imagine that? 21:26 Yeah. No, that's a real barrier. Yes. And this is a huge gap between software developers working in open source and creating and helping each other building things. Yes. Yes. 21:37 But open source and education is just not normal. Library is open source, right? But that's the only thing. Yes. 21:46 I mean, the, uh, uh, initially about when we were starting out, there was a lot of OERs, open educational resources. But, you know- Running out of those... they're not that, that... 21:57 Wikipedia, I mean, literally was around. But the problem with Wikipedia was they were not foundational. They couldn't be, you know, kind of targeted to young kids. 22:08 Kids could just go to Wikipedia and, uh, a lot of the, um, you know, things in, uh, online couldn't be used because of they weren't at the right level for a kid, and you didn't know the pre-knowledge of that kid. 22:23 Another non-negotiable is you have to figure out what do they know? And that, that's one problem w- online that became very, uh, important, the cold start problem, which was about, you know, 22:37 how do we give this child something which will help them or which will make them wanna learn because they understand what you're saying? If it's too hard, too easy, they're gonna lose interest. Yeah. 22:51 If we go back to philanthropy, it just... Yeah, how to make it more impactful on the systems that are hard to penetrate in that sense? You know, there's lots of ways to do that. 23:04 If it's all about charging to keep yourselves going, part of that is you have to stop and say... 'Cause education in the US was getting very expensive. 23:15 And, and the things you have to say is, "Look, this is one of those things that we cannot, you know, kind of waste or cannot just ignore. This is a public's right- Yep... human rights kind of issue." 23:30 And without that, society itself... 23:32 I mean, if you think about if the US wants to keep its, its, uh, advance, what, where they're going, they're gonna have to make sure that the next generations coming up are equipped- Mm-hmm... to take over. 23:47 Do you see this happening? Well, the, the most kids... There are very few kids who are qui- who are asking those, the why question, the whatever curiosity question, which lead to learning, right? 24:00 I think it could happen very fast because in China, India, kids are hungry. And a lot of the other countries, kids are hungry because they understand education is one way, is the way to get out of 24:16 their, you know, wherever they were born. Imagine if you are with a poor, uh, family. Yep. Your biggest need is to make sure you kind of contribute to putting food on the table, shelter, et cetera. 24:32 Your parents can't help you to study. We can. That's our job. My big hope is that AI can Really accelerate that. Do you- Yeah... do you feel the same way? Yes, because like I said earlier, that AI gives us a deeper... 24:48 So we can do deep knowledge tracing- Mm-hmm... with AI, with the algorithms, you know, all the technologies we build together on our back end. 24:58 We can deeply know where exactly this child was not or misunderstood or m- mis- under- there was a misconception, um, and help them bring them up faster. 25:10 The system, education system today says every year every child should be able to make up whatever they lost for how many years back to what they're learning today, and that's almost impossible given how much you might have missed, especially kids who are not, you know, who don't have the privilege of having a mentor or a helper. 25:34 That's why our Flexi is 24 by 7. It's a mentor for them. Yeah. Well, the mentor and parent is also very important for children. 25:46 And if I can ask you a little more personal question because you really strike me as a grounded and modest- Yeah... person. And with all the resources you have- Mm-hmm... 25:56 how do you raise children to stay really humble and grateful? What are your core values? 26:03 For me, I think the core values are about caring for everyone, animals, human beings, and that's something all my kids have inherited. Mm-hmm. 26:13 That's one thing, and I don't think you have to flaunt if you're from privileged family. You should be... 26:20 You know, we, we have encouraged our children to not think about going to fancy places of popular places, but go out, like, we'd go to the Amazons or we'd go to Galapagos or... Yeah, those are still pretty... 26:36 But, but give them- Exploring... exploring. You know, when they were in, um, middle school, because, um, uh, we collected a few kids whose parents would- wanted to do the same thing, 26:51 we asked a teacher if she would go with the kids, and we were willing to fund this for a- any of the kids who wanted that. The school did that but at a different level. 27:01 We had her take them to South America and live in a village with... and help build toilets, help build whatever needed to be built, you know, to learn to see what most normal people- What's the life like. Yeah. Yeah. 27:19 And those are the c- core values you have to have. Help anyone you can. Uh-huh. Whether it's money, whether it's mentoring. I wish... 27:28 There are so many women in this world who get educated, then get married, 'cause that's what's expected, have children, take care of the children, but their expertise or their talent is not going anywhere. 27:44 What if we asked all of them to kind of, kind of adopt kids around them and teach them to be their personal assistant- Mm... to be on call if they didn't have computers, holding hands across the world for the children? 28:02 I think it could go a long ways. That's true. I see that a lot in our community, but we live in a small town- Yes... where people can do that. 28:10 I think it's just much harder in the bigger cities, and much more people live in the bigger cities now. Yeah, but you know something, if you have young kids, the schools. 28:20 So one of the things about the school in this, the school that my kids went to, they had lots of volunteer parents- Mm-hmm... who were, you know, uh, very experts at different things- Mm... 28:32 and they would bring their expertise into the classroom and help the teacher. You can do that. Volunteer your time. So yeah, it's more about showing than telling. Yes. Show up. All of us have to show up. 28:45 All of us have to show up, yes. And, you know, finally, it's just like attention is what you need was what taught the machines, basically. 28:54 No one's taught that, but the, the transformers and atten- kind of the next word kind were the most. But now in this AI world, we need to have intention and judgment. Intention is what we need. 29:11 Why am I learning this? You know, it just can't be all, "Let the AI learn." That's what I would say. What is the most exciting for you in this new AI world right now? Two things. 29:26 Number one, going to be a grandmother for the first time. Oh, congratulations. [laughs] Right. No, no, no, but, uh, no, that's not the point. The point is that humanity will continue. Mm-hmm. 29:37 Even though we have limited time, but we will... Selfish genes will continue. Humans will continue. And number two is that I think if we use AI as augmented intelligence that's helping us, we can do wonders. 29:55 Don't be afraid. That excites me. I 100% agree with you. But if... Exciting part is important, but what are your concerns? I think it's also very important to discuss that. 30:06 Oh, you know, fear is a huge concern, and very often fear is founded in reality, which means this, as long as humans have existed, there have been fights. 30:21 We fight with each other for dominance, for whatever, right? And I think there's always gonna be players who want to shortcut the system. 30:30 Crypto came along and some people just, okay, the mi- you know, kind of Ponzi schemes and all that. Yep. Stealing other people's identity. Those are realities. Those are things that we're gonna have to deal with. 30:45 But again, like I said, no system, no technology comes... Even a baby doesn't come ready-made. And so when you think about that, we have to honestly get in there and keep making it better. 30:59 But then make it available to everyone and don't be the hound that takes all the spoils [laughs] from the whatever people that... Anyways, these are some of the things that kind of scare me. Well, thank you very much. 31:10 And I always ask this question at the end of the interview about a book that significantly influenced you, and it might be a recent book or something that impressed you in your childhood. There are a few books. 31:24 One of the books that really... I learned to read English by reading the Seuss books, you know, um- Yep, Dr. Seuss... Cat in the Hat and all that, right? This is still very important. 31:35 I learned to read through Archie Comics, English, right? As I grew up, The Hobbit was very influential- It's a great story... in my thinking. 31:45 And then Man's Search for Meaning, Viktor Finkel- Frankl's book really impacted me a lot. How? 31:53 Because it was about concentration camps, how people kind of kept their sanity through these, um, tough times. 32:03 We all have our things that we need to deal with, and I think that kind of really hit home to me about how might we even tell our kids how, how tough times may be hard, but how do you get through those? 32:20 You know, those are some of the things I've been thinking about. Yeah, and people are also very adaptable, and you want them to adapt to something better. Yeah, and I, I love music, so I'm not... 32:32 You know, the music's something like, you know, Hamilton. Have you- Mm-hmm... ever heard that play? The musical? Yes. And the song, I'm not gonna lo- lose my sh- my shot, you know, by- Mm-hmm... who wrote Hamilton. 32:46 I always forgot his name. Yeah, I know who you mean. Yeah, you can look it up. Yeah. [laughs] I still... I'm not giving away my shot. I am not throwing away my shot. I am not throwing away my shot. 32:58 Hey, yo, I'm just like my country. I'm young, scrappy, and hungry. And I'm not throwing away my shot. I'm going to keep taking the shot and making sure... And, and, and you know the idea people think they don't... 33:10 You have to learn lifelong. I was 50 when I went back to s- college to get another degree. Oh, wow. In education? Yes. I was 52 when I started working on CK-12. Mm-hmm. I'm 70 plus, and actually don't write all this. 33:26 [laughs] Actually, why not? I'm 70 now. What's next for you to learn? Oh, so much. So much. 33:34 [laughs] I'm trying to learn, you know, some of the things, like I don't think about that I need to learn is they talk about economics and the economics of, you know, with AI coming in. 33:46 I'm not worried about this new technology changing, giving there are no jobs. But what I think every time technologies come along, new medias come along, it's opened up doors for other jobs- Absolutely... 34:01 and taken away jobs that don't need human beings to suffer through. AI is gonna do some of that. That- That's going to be a transformation... that is. The question people are afraid of, you will have to re-skill. 34:14 That's what we want our children to learn, re-skill. Broad learn, broad learn quick so that you can... And but truly learn and apply it. Wonderful. Let's do that. Okay. Thank you. Okay. Thank you so much. 34:29 [outro music]