Sal Khan, Founder and CEO of Khan Academy, discusses the impact of artificial intelligence on education.
Hosts: Carol Massar and Tim Stenovec. Producer: Paul Brennan.
You're listening to Bloomberg Business Week with Carol Messer and Tim Stenavic on Bloomberg Radio. Children, Well did slowly go and all right? Teach your children? Well right, yes, we hope. I know my child who's not really a child anymore. The music that they pick is just so on point. So thank you to the team who does that. Yes, why yeah, Crosby Stills love it. Um. When it comes to teaching kids, sometimes they need an assist. And I have to say my daughter over the years has definitely tapped into Cohn Academy. Yeah, business school exactly. I mean it got me through business school. I'm really excited to have with a s con founder and CEO of CON Academy. He joins us on Zoom for you from Mountain View, California. Uh So, I gotta tell you, I promised myself when I finally did get to interview you this many years after business school, I would thank you for the videos that you did about discounted cash flows and interest rates and all the things that I worked on in my personal finance and accounting classes because it was your voice that helped guide me through business school. Uh. And it's interesting because this Carol wasn't available when I was an undergraduate. We weren't even talking about YouTube back then. Very please to go back with the cell con founder and CEO of khn Academy joining us from Mountain View Soul. How are you. I'm doing well. Thanks for having me, you know, take us back to two thou eight, because it's funny and thinking about this interview that we're doing right now. You know, we think about online education in the way that it really gained so much attention in the early part of during lockdowns and when kids were sent home. You've been working on con Academy for a dozen years by the time the pandemic hit. Take us back there and where we are now in some ways earlier. It all started back actually in two thousand four, when I was tutoring some cousins remotely when I was in Boston, they were in New Orleans. I started writing software for them. Friends suggest that I make videos for them on YouTube. It took on a life of our own, and it was in two thousand and eight that I actually set up this, this family project as a not for profit mission free world class education for anyone anywhere, and called it kN Academy. And if we compare that time to where we are now, I think now people actually take it for granted that, of course, on demand video is going to be really useful, uh for for learning. Of course, students should be able to have self paced practice. Um. Of course, it's not the best use of class time for a teacher to just lecture at students and students to sit passively and listen. Because students have access to on demand video, perhaps it might be a good idea to allow students to learn at their own time and pace. But if you go back to two thousand and eight, none of that was mainstream. People thought things like YouTube and maybe the Internet generally for the most part, was as a toy or or maybe a distraction. But now we realize that it's it's central to the education. I think the other really good thing that happened is, obviously when when we set out as a not for profit in the name of trying to level the playing field, it's it's predicated on the idea that everyone eventually will have access to the internet. In two thousand and eight, we weren't anywhere close to that. We're still aren't at perfection, but in school environments we've gone a long way in the US globally that there's still a lot more work to be done, and I think we the pandemic. One of the silver linings is it it helped accelerate a lot of internet and tech adoption, not just in schools but at home as well. How do you think it's you know, the impact it's had on how kids learn um or helped compliment, supplement you know, classroom time. Yeah, well, I think you know, everything is is a double edged sword. We all know about all of the things on the Internet that maybe aren't the most productive things for kids, but in terms of when they are productive on it, whether you're talking about kids in elementary school or middle school, all the way to med students, it's now become pretty mainstream that students are going to vote with their feet if there's someone who's just going to lecture at them and the students have an option, and they won't show up anymore. If you if you visit most med schools, the actual electoral halls are pretty empty. The med students, who are very diligent students are usually going to watch the lecture later on at double speed. UM. So I think there's an expectation of Some people say, oh, kids these days have less of an attention span. I don't think they have less of an attention span necessarily. It's more that they have more options and they have higher expectations. When we were kids, we're like, Okay, we gotta sit in the classroom and just pretend to pay attention and look at the clock, and so we're gonna play that game game. But but now now you don't have to do that. They have they have other things. I think they know that if they don't understand something in class. Back when we were kids, you'd have to either somehow struggle with a textbook, or if you're lucky enough to have a family member help you, or or had enough resources to get a tutor. And now we're talking about very few people. Now people know, yeah, I can go to con academy, I can ask the questions, I can find other I can do practice problems. I can make sure that I understand what I'm doing. I can go to YouTube and find other resources. So I think kids are actually able to advocate for themselves a lot better because they don't. They're not willing to just put up with only one modality. Hey, given all the work that you've done over the past gosh, I mean it's close to twenty years at this point. So I'm I'm wondering, you know, how, how how it changes the way you think about education with regard to your own kids. Yeah, even back in two thousand kind of candemy really got on on folks radar. In two thousand nine, two thousand ten, two thousand eleven, I gave a ted talk whereas, you know, let's let's reimagine education, make things more personalized. We shouldn't have kids moving forward lockstaff if you don't. If you only got a seventy percent on the test, you should always have that opportunity incentive to make that an eight percent, make that nine. And I wrote a book, One World school House, about how how education could change, how you could even create a school of the future. And then in my oldest I now have three kids, my oldest was just entering kindergarten age, and I said, hey, I don't want to be a hypocrite. Um, I don't want to preach all of these things about mastery, learning, personalization and then not have my own kids do that. And not only that, but I actually did and continue to believe that that is the most powerful way to learn. So we started a school, conn Lab School uh where now all of my kids go. It's out here in northern California. And what we're seeing is when you have a school that's really students centered, where you know we have. The one edict we have is there's no lectures at the school. Everything has to be active. Students can learn their own time and pace. But once again, they're not in isolation. They get support from their peers, they get support from the faculty. When human beings are in the room together, they either have Socratic dialogue, or they're doing projects, or they're collaborating, or they're tutoring each other. We have a motto everyone's a student, everyone's a teacher, and we're seeing we we've had our first few graduating classes in the last couple of years, and I don't want to jinx it, but the kids are doing frankly better than you would expect from a traditional environment. We just looked at our lower school, which is you know, grades K through six, K through five, and our kids are seeing about two grade levels of math a year. We just launched uh CON World School with Arizona State University, which is an online high school, and they they're using similar modalities but in an online framework. And they just saw on the first semester five times the expected growth in math, and I think it was three times expected growth in reading. And but once again, it's just because the kids are able to learn at their own, their own pace, and because they're not learning to be passive and they're learning to have more agency over their learning. They're more entrepreneurial, they have more of a growth mindset, they're willing to take risks. We think they're feeling a lot less of the stress and anxiety that we know as a bit of an epidemic right now amongst young people. So there's not a huge history with con Lab School because you only started it a few years ago. But where are you in terms of college placement and how you've been able to, uh see how these kids have done longitudinally. Yeah, and as I said, the school has been around for about eight years. We had this current guarding graduating class is going to be our our third But you know, this is a school that does not have traditional letter grades we measure, but at any point students can improve their grade, but out of you know, I remember our first graduating class of nine students they went to and I'm very sensitive because so many folks index on where kids go to school and all of that. But our students ended up at some of the top places in the world. Um, you know, I don't want to get to brand conscious about where they went, but they went to let me just say, very very good, very good places. And they're even more importantly, we we stay in touch with these students and they're really thriving in these environments because frankly, the college environment is all about self based learning. The college doesn't pretend that way, but really that's what it's all about. No, you're absolutely right. You get what you put into it, essentially when you go to college. Hey, so, sal we're looking at your school, um, collapse school. Can you scale that? You want to scale it even further? It's in Mountain View, right, but can you go further with it or do you want to? Yeah? The whole idea and I always tell the team there that is not just to start another school in Silicon Valley for my kids and other folks in this area. Obviously you have to serve the kids in the community. Well, but it's really to show there's another way of doing things and then essentially sharing with the world how to do that so it can it can scale. So one of the first ways of scaling is by starting CON World School with Arizona, a s U that's actually free to any student in Arizona because it has a state charter there, and then it's we we hope reasonable cost outside and we're working on more states, so you really can get a free world class education if you're if you're in the states that that supports something like this. So that's one skilling mechanism, and we are seeing folks form essentially hybrid pods where they're getting the benefits of in person sports, etcetera. But then they're able to use the World School UH for their socratic dialogue and their seminars and you know, being able to meet their advisors and things like that. We also are starting more CON Lab schools. It's likely we might start one in the Midwest UM and so I think we are at the moment. I won't ever say we have nailed it. It's a lab school. We're constantly iterating, but I think we're starting to scale it and that could take multiple forms. It could take an online school. It could it could be a share the curriculum with other folks. What what, however, we can get it out there, Salgion. The numbers work. It's it's really tough to be self sufficient as a school. It's a it's a private school, so there's you know, tuition. Um. But I mean I went to a school that my friend's mom started for second grade and the school was only around for about ten years. They just couldn't keep it going. It was tough. Um, are the numbers working. The numbers are working. And what I've always given the charge to the school that we should be able to be in terms of cost per student per year, cost competitive with um some public school districts, and they're huge variation. If you look at at the high end, a New York City public school spends about forty dollars per student per year. A lot of you know Cambridge, Boston, Massachusetts, they are mid twenties, even approaching thirty dollars per year. But then you have other places, including high cost of living places like California, where depending on the district, you're really looking at more of like ten to fifteen thousand dollars per year. Right now, if if you were to hold all things equal, we have we have to pay a lot in real estate expense because we're renting out a couple of campuses. But if our real estate expense we're similar to what a traditional public school was spending, where we we have our cost per student down to about twenty dollars per year UM. And and so if you include the real estate that it's it's approaching a higher number. It's you know, it's probably closer to twenty eight thousand, but UM. To answer your question, we we intentionally are keeping it much more accessible than the surrounding market would support. These numbers don't sound like small numbers to anyone anywhere, but especially but if you live in the Bay Area, if you live in New York, you'll know that a lot of the top independent schools will easily charge you fifty or sixty dollars a year UM. And you know, we're we're we're pretty focused on showing this model can scale and as an as is accessible and also has evidence that isn't just another independent school. It's the kids are growing faster and happier. You know, in doing this, what been the easy part, what's been the difficult parts of doing everything, including school or con academy or the whole everything from the get go. Like I think about when you started, and I know I remember, you know, we talked about Bill Gates being you know, a supporter of it. You had some really high profile people, and I'm sure that helped um, you know, doing the Ted talk. But I'm just wondering, as you've done this, what's been the easy part, what's the difficult part. Well, the easy part is this is something that I've always been drawn to, even when you know my past, my past careers in tech and just been an analyst at a hedge fund, which I both careers I enjoyed, but I always was drawn. That's why I started tutoring my cousins. That's why I started writing software for them that was focused on learning. And that's why I started making videos because I enjoyed making these these education videos. So that was always the easy part. And whenever I want to recharge, even today, that's the kind of stuff that I try to really work on. Let me make some videos today. The hard part, I would say, the early days, the hard part is convincing people to take you seriously, especially you know now it's a somewhat mainstream thing to be a YouTube influence or whatever else. Back in two thousand and six, two thousand seven, two thousand and eight, it was very um let's call it. No one really took you seriously. And then once we started to get off the ground and people took note. I think when we said no, we just don't want to be a YouTube channel, We don't want to be another for profit education company. We actually want to create an institution for the world that actually has a chance of being the safety net education system for the world that can reach billions of people. I think back then some people might not take that so seriously. For a guy operating out of a walk in closet, UM, I think you fast forward and yes, you know folks like Bill Gates and the Doers, and you know there's so many people, um that you know I grew up reading about who who have now become some of our supporters. But whenever I say those names are mind folks that we still need their help, so please fo to kind of academy. But um, I think now what's gotten easier is when when I take on a project. Let's say we started school. We're starting this online high school, or we started another not for profit called Schoolhouse dot World, which gives free tutoring. They way it's able to do is to leverage volunteership. These types of things would have been very hard when I was just a guy in a closet back in two thousand eight or two thousand nine. But now people are willing to take it seriously, They're willing to partner. Uh. You know, the fact that even volunteers show up to Schoolhouse dot World and to tutor other people is because there's some trust and belief in what we're trying to pull off. But even today, I think I think it's a lot of a lot of what I do is trying to get people to believe I really I realized that that's like the believe both in our own organization. Like hey, hey hey, folks, we're literally trying to move the dial for the world. And I know you can get cynical about that, but it's doable. So how do I want to talk AI? Because you know, we talked a little bit about BuzzFeed and rising today as a result of excuse me, as a result of a partnership that announced yesterday with open Ai C three dot Ai share surging more than seventeen today. Um, how do you use AI and a vacation? Well, there there's you know, there's a lot of what we already know is going on, and then there's a lot of potential. Even before these large language models like chat, GPT and all of that, we've always thought, hey, AI could be used to this is it's being used in for profit companies to recommend the right add to you. In an education setting, maybe it can be used to recommend the right content to you. And we've used variations of that in the past at Kon Academy. I think what's exciting about these large language models is the potential for them to start to act like a Socratic tutor, for them to intro to introduce modalities that might not have existed before. I know people are really worried about this doing essays for students, but maybe we can embrace that and say, well, maybe it can help the student do something more ambitious. Maybe it can help give feedback that Traditionally, when you write an essay, at best, it might take you a week or two to get the feedback, and then if even if you make it better, you may or might not get a chance to to get feedback on that. Now you could get instant feedback. Uh. I did a little experiment with my daughter where we used a large language model to co write a story, but then she had a chance to talk to one of the characters. That's something that just seems like science fiction, but it's now doable. Or we're at the cusp of of making these things doable. So if you imagine, I think we're in the next year, we're going to see layers on top of tools like con academy where you might have a tutor and AI tutor. Now, once again, I don't think this replaces the human, but it just gives more leverage to the human teacher, to the human parent, uh, to be able to be able to do more. It's like moving beyond when you know math right. It used to be what's what's the answer? Did you get the answer right? And then it moved to wait, what was your thinking and show us the work? And okay, you know that was more important than really kind of the final to some extent or as important to getting the answer right. Hey, just quickly, one last thing. Um just got about forty seconds left her What do you think about the a C T S and S A T S. A lot of it went away during the pandemic. Good move of like do we have to move away from that stuff? Well, those tests is perfect. But I always tell people if you don't like standardized tests, what part do not like the standardized or the test? And if you need to have some way of of of understanding, and especially now that you have chat GPT that can write your college essays, this is actually the fairest mechanism by which to be able to see who's college ready brilliant we have not. That is just an interesting way of thinking about this and I don't like test part. To answer your sal thank you so much, good luck. Such an incredible organization that you've created. Um and as we said, what almost twenty years in, it's pretty remarkable. Sal Con He is founder and chief executive officer of con Academy, joining us via zoom from Mountain View, California. Just google them you can find out all the incredible work that they are. Yeah, it's incredible. For a long time. More than fifty languages used in more than a hundred and ninety countries, a hundred and forty five million registered users,