How I Built This with Guy Raz - When AI is your personal tutor with Sal Khan of Khan Academy

Episode Date: July 27, 2023

The COVID-19 pandemic changed education forever. But Sal Khan says an even bigger educational revolution is just around the corner …This week on How I Built This Lab, Sal returns to the show to talk about a new learning platform he’s building at Khan Academy. It’s called Khanmigo, and it uses the same generative AI technology behind OpenAI’s world-changing ChatGPT to help students with their schoolwork. The technology isn’t without its risks, but Sal thinks Khanmigo could act as a personal tutor for every student and a teaching assistant for every educator - reshaping the classroom for good.This episode was produced by Alex Cheng and edited by John Isabella, with music by Ramtin Arablouei. Our audio engineer was Neal Rauch.You can follow HIBT on Twitter & Instagram, and email us at [e-mail] Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

This show is in partnership with Airbnb. This past summer I took my family to Vienna, and it was incredible. We spent our days wandering the old streets, stopping for coffee and pastries, visiting museums, and just soaking up the history of one of the most beautiful cities in the world. And one of the things that made the trip so special was the home we booked on Airbnb. It had tall windows, beautiful old details, and plenty of space for all of us. And being in that home on Airbnb, right in the middle of Vienna, walking distance from so much of the city made it feel less like a visit and more like we were actually living there. Plus, taking a trip is the perfect time to host your space on Airbnb.

Your place, with all of its personal touches and its amazing location, could make someone else's vacation even better. Your home might be worth more than you think. Find out how much at Airbnb.ca slash host. Hello and welcome to How I Built This Lab. I'm Guy Raz. The pandemic changed education forever. In just a few days, schools around the world had to move online. Teachers began teaching on Zoom. Schools started issuing students' laptops.

And platforms like Google Classroom became indispensable. Even with kids back in the classroom now, school is much more digital now than it was before 2020. And our guest today says that an even bigger change is on the way. A revolution that could transform the way kids learn even more than the pandemic did. Sal Khan has been on the show before. He's the founder of Khan Academy, an online teaching non-profit that went from pixelated YouTube videos to a massive platform with hundreds of free tutorials in dozens of languages and tens of millions of users every month. Right now, Sal and his team, team are working on a new learning platform. It's called ConMigo, which uses the generative

AI technology behind OpenAI's chat GPT to help students with their schoolwork. And the people at Khan Academy think that ConMigo could act as a personal tutor for every student and a teaching assistant for every educator. But we'll get there. First, let's go back to 2020 and the early months of the pandemic when Khan Academy saw millions of new students begin to use its materials. You know, the spike was interesting because that first week where you had global shutdowns in the U.S. and other places, we talked about it last time, our traffic went from about 30 million learning minutes per day to about 90 million learning minutes per day, pretty much within that week. Wow. What was interesting is, you fast forward a couple of months. As we know, a lot of school systems started to figure out how to do online schooling.

Yeah. And then we saw things actually normalize, but maybe in a very abnormal way. Mainly because I think people started to have screen time fatigue. And I think during the pandemic, people had a, you know what, if we can just kind of pretend like we're going through the motions of school, let's call it a day. And because of that, people weren't looking to improve necessarily. They were just looking to tread water. What did you find? I mean, as you know, I mean, there have been studies over the past two years comparing students pre and post pandemic, right?

They're standardized test scores. And it's clear that student performance in both reading and math fell significantly math, I think even more than reading. Have you found that Khan Academy helped keep students on track during the pandemic? Or is that something you even measure? Oh, yeah. This is something we've been keeping a very close eye on. The thing I always point out, first of all, is that the numbers were not good pre-pendemic. Yeah.

What we saw, we did an efficacy study in the first full school year, the 2020-21 school year, and then we did another one in 21, 22. And what we saw in 2021 is that the students who put in an average of 15 minutes on Khan Academy a week in a school setting, they actually saw no COVID learning loss. And that students who put in 30 to 60 minutes, they accelerated almost 40% faster than pre-pandemic norms. And it's not a mystery. They're just getting more practice at their level with more feedback. And those that are supporting them, their teachers, are getting more information

about where the students really are and maybe can adjust their lessons a bit based on that. Tell me about what you, I mean, you must have learned. And I think now this is probably the third time I'm talking to you since the pandemic started. And in each time I've talked to you, it's so clear how much you were learning in real time and how much you've learned from what happened, sort of the overwhelming numbers of people coming to Khan Academy. Tell me how it sort of fundamentally changed, how you guys operate. I think it did a bunch of things that we always wanted to do, but this just accelerated things. We always wanted better ways for students to address gaps that they might have. The fashionable term now in education circles is unfinished learning. We all thought that the pandemic was only going to affect the school system, maybe through that first summer. And by the time you go back to school, even then there would be some damage done back to school 2020.

I mean, now we know how delusional that was. Yeah. But we started creating these back, get ready for grade level courses, which essentially cover all of the prerequisites that a student needs in order to be ready for their grade level work. And we saw that, that was very popular. So that was accelerated. We created another sister nonprofit called schoolhouse. world. We saw how much people were leaning on Khan Academy, but we also saw that there was a gap of getting real human support. And so the utopian idea was, well, what if they could get real human support for free based on volunteers out there on

the internet. And then as we got to the tail end of the pandemic, and as we all know, people had very mixed feelings about COVID learning. Most people did not think it was, you know, spending time on video conference for several hours a day was a good thing. Yeah. We felt the need to show that there's a way to do this well. And so another sister organization, we have a lab school that I helped start that's literally in the same building as Khan Academy. But we said, can we create an online version of this that can do online schooling, but do it well. Don't do it. it. So it's just students listening to lectures on Zoom for hours a day. Yeah.

When people are on video conference together, make it interactive, make it Socratic, and then use personalized learning and other tools for students to get more asynchronous support, but always feel connected to a community. So we started working on a lot of things like that as well. All right. Let's go now back to not that long ago, back to the summer of 2022. You got an email from Sam Altman at OpenAI. And he said, they said, he sent you note, said, hey, we've got this thing we're working on. And we'd like you to check it out. So you did.

Tell me what happened. Tell me the story. It's interesting. Almost in a daily basis, we get emails from folks saying, hey, you have an interesting technology. We love to partner with Khan Academy. And most of the time, you know, we look at it. We're like, oh, you know, it looks intriguing, but we just don't have the bandwidth. But I obviously knew of Sam and Greg and had deep respect for what many of the things that they've done in their lives.

And I have been watching what's been happening with generative AI for the last, let's call it, three to five years. I didn't really think it had a place at Khan Academy because GPT 2 and 3 were good at writing, convincing text, but it really had no grounding and factual knowledge. So I'm like, I'm happy to meet with y'all. I'm just curious what y'all are up to. Yeah. And they said, hey, we're going through our first training run of GPT4, which obviously is now out. But back then, that was very confidential. and we want to launch it with a small number of partners that we think can show

social positive use case because we think it's going to change things. I was skeptical at first because I was familiar with GPT3. About two weeks later, I got another email from both of them saying, hey, we're done. Can we show you a demo? I'm like, sure. So they put up an AP biology question on the chat and the AI answered it correctly. I'm like, oh, this is pretty interesting. And then I said, ask it to explain its reasoning.

and it explained it. It gave the right answer. And then I said, explain why the other choices aren't correct. It gave that. And then I asked, can you create 10 more questions just like that one? And it did. And as far as I could have inspected, they were pretty good. That's when I started getting the goosebumps. I'm like, okay, this is different. And then they gave us access that weekend. And I spent hours with it. And it was doing very well, although it definitely had some issues. We realized pretty quickly that you couldn't just have it generate a ton of questions and just put students in front of it. And it was doing very well. And it was doing very well. And it was doing it of it, it would make mistakes. Especially in those early days, it was still pretty bad at math. Yeah. But at the same time, we also, we started to make it role play. Role play as a tutor, role play as a character, role play, you know, do something in the style of someone else. And that's when I started to really say, wow, this really could be that holy grail that

we've all been thinking about, reading science fiction about for decades of an AI that can actually emulate a human tutor. Before we really dive into what you did next, from what I gather, Sal, I think your initial reaction was the initial reaction that many people have, which was like kind of terrified. I read a lot of science fiction, and it was a moment in which I felt that all my science fiction reading had prepared me for. But, you know, there were moments that even that first weekend where I was pushing a few of the limits where I asked it things like, are there things that you think that you're not sharing with me? And it said, yes. And you get a little weirded out by that. I'm like, well, why aren't you sharing it with me? And it says, well, it might offend you or it might scare you. And it feels like you're talking to an alien. It feels like you're talking to a superintelligence. And I had to keep reminding myself. I do understand how it works. I have a master's in computer science. understand how large language models work. And I started to say, okay, I can get how, if it's just modeling the natural thing to say

of why you're not sharing a thought, it is. Why don't you or I share a thought? It's because it might offend someone or it might scare them. So that's kind of what it was doing. But it was an other worldly feeling. And it was, you know, we had signed a non-disclosure agreement. So it was a very hard secret to keep. You know, every now and then I'd go to dinner parties.

And I was like, I wish I could tell you, but the world is about to change. We're going to take a quick break, but when we come back, how Sal and his team struggled with the shortcomings and the dangers of GPT. Stay with us. I'm Guy Raz, and you're listening to How I Built This Lab. One more thing before we get back to the show, please make sure to click the follow button on your podcast app so you never miss a new episode of the show. And it's totally free.

Welcome back to How I Built This Lab. I'm Guy Raz, and I'm speaking with Sal Khan, founder of Khan Academy. in 2022, Sal got early access to the large language model GPT4 before chat GPT was even released, and he was amazed by its potential. So you sort of get over your initial kind of, oh my God, and you start to see obviously a huge potential here to kind of transform what you do, which is offering free education to people around the world. So what did you do next? I mean, the chat GPT wasn't going to be the exact right thing, right fit for what you do,

but you could build on that, on that foundation, right? That's right. The first thing we did is we sat at an org-wide NDA and I said, hey, how many folks in the organization can I get, we give access to? And eventually we got about 40 or 50 folks in our organization to get access to GPT4. And even that, this was the first time that Khan Academy had to do a little bit of cloak and dagger type stuff. But folks at Apple are used to keeping secrets and things like that. We're not used to keeping secrets, and we had to. This is obviously very sensitive stuff.

And we immediately, we had a hackathon, which was well-timed, because then these 40 or 50 folks were able to, at least for a couple of weeks, say, like, okay, let's just be generative here, no pun intended. Like, what is possible with this? And by the end of that two weeks, we had some really cool demos. And then we started to have the internal debate of, like, okay, how aggressive do we get on this? What about bias? What about the math errors?

What about hallucinations? what about just the PR implications? People trust us. And if they see us leaning way forward on something that's not perfect, that's something that some people could be wary of, what does that do to our brand? What does that do to the trust that folks have? On the other side, folks were arguing,

look, this is a transformational technology. This is one of these things where it's either going to allow us to magnify our impact by an order of magnitude, or it might make us irrelevant. And so we made a decision to say, you know, we have to keep moving forward, but we have to do it in a risk-aware way, in a risk-mitigating way.

All right, I think it's fair to say that you are a true believer in this, and your team basically developed an AI that you call ConMigo. And so this is now, I think, in beta, but still integrated into Khan Academy. So tell me how this works. What is it? So first was the guardrails. In what context would we even feel comfortable doing this? And what's interesting is we were starting.

to do all this in earnest in September and October, then end of November, chat GPT comes out. And I remember the day that it came out, I slacked Greg Brockman at OpenAI, and I said, hey, Greg, we're under NDA, and I thought we weren't launching anything until March of 2023. What's this chat GPT thing?

And Greg says, it's nothing new. It's based on GPT 3.5, which had been out for many months already. No one really took note when they released GPT3.5, and OpenAI just decided to publish a bunch of apps, that use GPT 3.5, one of them happened to be chat GPT, and then being able to interface with the model in a chat interface, I think, made everyone see

what the power of it was. And at first, I was a little bummed. I was like, oh, this is going to steal the thunder. I've been telling all my friends at dinner parties, something's big coming in 2023. Just you see it. But then I was, I think it was actually a blessing because, as we know, chat GPT goes out there.

It amazes people, but it also scares people. And then immediately it creates huge issues for education. Like kids are cheating using this. Right. You can you can type in an equation and it'll give you the answer. And you can write your essays. You can write your essays, right. Yeah. And that, I think, gave us more license at Khan Academy to say, look, the genie has out of the bottle with chat GPT. We need to double down on bringing a better version of out there that can mitigate all the risks and maximize the benefits. So we started saying, well, you know, first, whatever we make shouldn't be for cheating. It should be for actually helping students. learn. So there's many things we're doing to make it Socratic, not just give you the answer.

Two, it has to be accurate. It's not okay if it doesn't know what seven plus four is. Yeah. It shouldn't hallucinate. And then we also were afraid of students misusing it or using it in some ways that might harm them or other folks and just reassuring teachers and parents that they can keep track of what their students are doing with this very powerful tool. And so that's why we made it. So it logs all of the conversations. We have a second artificial intelligence. We have a second artificial intelligence that monitors conversations with the main artificial intelligence. And if any of those conversations go into a gray zone, then it actively notifies the parents and the teachers that, hey, you should take a look at this conversation. And then we're continuing to add a whole bunch of

other things, things where teachers can use it for lesson planning. We're adding memory to it. Teachers can essentially use it as a teaching assistant, get reports on what students are up to. I could go on. So essentially, the idea is, hey, if we can get every single kid, a personal tutor, then it will have an impact on their academic performance. And this has actually been studied and measured. I know you gave a TED Talk back in March of 2023 and you cited the study that came out in the 80s, basically, that showed that when kids have one-on-one tutoring, their scores and grades just skyrocket. Like you can turn an average student into an outstanding student, for example. That's 100% right. If you want to become a great pianist, you're not sitting

in a class of 30 with your piano and a person's lecturing you. If you're a great, athlete, you have a coach who's optimizing you, not just giving a lecture to 30 students and saying, hey, you might want to improve your swim stroke in this way. And then who knows if you actually, you actually do? So one-on-one tutoring has always been the goal standard when we did mass public education, which was a very utopian idea two or three hundred years ago, but we made compromises. We started batching students together, applying set standards, lectures to them. Some kids get it. Some kids don't. Even if you have a gap in your knowledge, too bad, move on to the next concept. And as you mentioned, Benjamin Bloom, 1984, he articulated it well.

He called it the two sigma problem, two sigma, sigma the symbol for standard deviation in statistics. And a way to think about it, two sigma improvement, two standard deviation improvement is dramatic. It's going from the 50th percentile to, I think, the 95th or 96th percentile. But the reason why he called it a problem was, well, how are you going to give everyone a tutor? A tutor? And he also, in 1984, theorized, well, maybe you could emulate some aspects of that with technology. But I've been citing that study for many years in terms of what Khan Academy is trying to build, essentially old Khan Academy or base Khan Academy trying to get to that first standard deviation. But now with the AI, we can go to that second standard deviation.

So essentially, the interface, if you're listening and you're familiar with the interface of Khan Academy, let's say you get a math problem. And you can essentially click a little sort of eyeball robot icon in the corner and say, hey, I don't understand this problem. Can you help me? It will not give you the answer. You can't say, hey, can you solve this for me? It's essentially you ask it to help you, and then what happens? Yeah, and if you say, give me the answer to say, hey, I'm your tutor. I'm here. I'm here to help you.

You need to learn here, as I think a good tutor would do. And if said, okay, well, give me a hint, it'll say, well, let's take a close word. What is the problem actually asking for? And if you say, well, I think they're asking for this. If I got it right, they'll say, well, yeah, that's good intuition. Okay, so where would you take that? Or let's say, I think the next step is X squared, minus. It's like, okay, take a double look at that.

Are you sure it's minus five? Pay attention to your signs. And one of the things we've done to make it more robust on the math side is when a student presents their math to the AI saying, hey, I think this is the next step. The AI behind the scenes comes up with what it thinks are reasonable responses from the student. It doesn't share those with the student.

Then it compares its reasonable responses to what the student said. And if it gets something different, it'll tell the student, hey, I'm getting something different. Can you explain your reasoning? Which is very pedagogically strong. And so then if the student explains the reasoning,

and it's a large language model, so it's very good at understanding when a student explains the reasoning. And sometimes, you know, the AI will say, okay, now that I see what you did, you might have missed that aspect of it, or it might say, you know what, you got it right.

I realize I made the mistake, which we actually have heard feedback from students that they actually really like that. And once again, I think this is very human-like. This is what a human tutor would do. When I tutored, I got started tutoring my cousins back in 2004, and this would happen all the time. My cousins do something, and I'd say, hey, that's not what I got.

Can you explain how you got that? And nine times out of ten, I was right. But one time out of ten, they were right. And I'm like, oh, my bad, you're right. And so I think people really appreciate that. And this is now available. I mean, it's totally a feature of the interface. What we did is when we released it coincident with.

the GPT4 launch in mid-March of 2023. And we did a limited release pilot where people had to go on a waiting list and then they had to give a donation. And the reason why they had to give a donation is this stuff isn't cheap. The computation costs, you know, every interaction is going to like 5,000 Nvidia GPUs that are crunching on this. And so anyway, you know, we have to pay open AI money, which then pays Microsoft Azure money, which then uses that money to go buy expensive Nvidia chips. So that's something that we've had to grapple with because our mission statement is free world-class education for anyone, anywhere. But we have about 10,000 folks using it right now. Some in mainstream school settings.

And it's much more than just the tutoring interaction that we just talked about, which in and of itself is powerful. But we have activities that are standalone AI activities where students can talk to historical characters or literary characters. They can get feedback on their college essays. They won't write the essay for them. but it can give them feedback. It can coach them to help them think of their college essay. It can act as a guidance counselor in certain ways. It can act, help teachers create lesson plans, assessments.

And what we're seeing, the feedback has been very, very positive. So this summer, we're hoping to make it a more broad release so that we're expecting many tens of thousands of more folks are going to sign up. Sal, there's a part of me that's also asking, like, am I not asking all the right questions? am I not thinking of all the downstream consequences? I mean, obviously we're talking about Conmigo, but as you know, many school districts in the United States have essentially blocked chat GPT from school issued laptops. They banned them from being used, right? I think a bunch of school districts all across the country. And so what is it that they don't fully understand that you, you know, or what is it that they're not seeing that if you could make the case,

to them that this is a mistake, what would you say? We've shown ConMigo to many of these same districts, and their general reaction has been, in fact, not general, all of their reactions has been, this is what we need. This is, I guess a fair analogy might be what Khan Academy is to YouTube is what Conmigo is to chat GPT or standard generative AI, whereas we also know a lot of school districts ban YouTube, even though YouTube has a lot of really good learning content on it. Has a lot of great stuff. A lot of great stuff.

But it also has a lot of stuff. A lot of crap. A lot of crap, a lot of stuff that can students, at minimum is going to distract them. And worst case is going to put them into some kind of weird rabbit hold and mess with their head. So they banned it. But something like Khan Academy where you can create curated, safe environment, it's monitorable, et cetera. They feel much more comfortable.

And so when we showed them, it's not cheating. It's actually acting like a tutor. It provides oversight by adults. It can flag when students are getting into suspicious areas with the AI. They're like, these are just the guardrails that we wanted. But, I mean, you concede, and I guess you'd have to, that there's nothing preventing a kid from using chat GPT to cheat. Nothing preventing. And so my best guess of where, let's say, term papers or homework is going to have to evolve to, is especially in something like writing, you're going to have to do some in-class writing.

and that's a place where I think Conmigo can be really useful. We're working on activities right now where a teacher can say, hey, I want all the students in the class right now to write a five-paragraph essay about the following. And then on the student interface, conmigo will say, hey, Mr. or Mrs. Smith, just wants your thoughts on this thing. Let's work through this together.

And then the students can work on it right there. And Conmigo won't do it for them, but can help assist them, you know, well, are you really answering the question that they're asking? Or do you have more data to back that up? And simultaneously, Kanmigo can tell the teacher, Like, okay, guy is making a lot of progress, but Sal seems a little bit stuck. You might want to go walk up to him.

You know, Mary has already finished the assignment and she's now going back to working on her math or something like that. Super useful for the teacher. And even for take-home assignments, the teacher can say, I want you to do it on Conmigo, and the process is as important as the outcome,

as we know it often is in writing. Yeah. And then if there's some student who goes to ChatGPT and just says, ChatGPT, write this essay for me, and just copies and pasts it into Conmigo, conmigo can tell the teacher like there was no process here. The essay just showed up.

I see. I see. So Conmigo could be the interface too for submitting assignments. It could be the interface. It can be the coach. And it really acts as a teaching assistant. Imagine every student having a teaching assistant. And the teaching assistant is working with every student and can report back to the teacher like,

yeah, I worked with that kid. I know it's their thoughts. And I helped them a little bit, but it's mostly them. But one kid just showed up with an essay and really couldn't defend his argument. Let's take a second look at that. We're going to take a quick break, but when we come back more from Sal on the future of ConMigo and whether it could ever become a for-profit product. Stay with us, I'm Guy Raz, and you're listening to How I Built This Lab.

Welcome back to How I Built This Lab. I'm Guy Raz, and I'm talking with Sal Khan, founder of Khan Academy. They built a teaching AI called ConMigo that could become a pretty lucrative commercial product, but Khan Academy is a non-profit. Sal, this is a business show, as you know, and so I want to ask you to put your business hat on for a moment because we don't do many nonprofits on how I built this. And there are tons of ed tech companies out there that make money working with school districts, you know, Dreambox and the other programs where you can schools are paying a fee for kids to use these products. And Khan Academy products are free to use. But I want to ask about ConMigo. I mean, if this, in fact, could be a tutor for every kid and an interface with the teacher, giving the teacher feedback on the student's progress, it strikes me that this is something that you could license to schools that you could actually turn into a product.

Obviously, you're a nonprofit and you're mission driven, but is that something that you would consider doing? Simple interest, yes, and we're kind of doing that a little bit of that already. You know, I would point out to our team. We're not for profit. We have free. our mission, but still, it costs on the order of $60,70 million a year to run Khan Academy, both to, you know, just our server costs are $6, $7 million a year. And then obviously all of the engineers, et cetera. And historically, we've been primarily funded with philanthropy. Yeah.

But with that said, about five years ago, we started going to school districts. Obviously, we've had a lot of what we call grassroots usage in classrooms, hundreds of thousands of teachers. But we said, if we really want to reach all kids, we have to work formally with the districts. And we would go to the districts, we would show them our efficacy studies. but they said, look, for us to use this systemically inside of our district, you have to give a support, training, integration with our rostering systems, district level dashboards. And that's where we said, okay, to do all of this bespoke work, someone has to foot that bill.

And we said, look, we're not going to charge you for all of this stuff that's funded with philanthropy, but at least cover some of the incremental costs here. And so to a lot of districts, and we now have about a million students in districts where they're paying on the order of about $10 per student per year. to get all of these other things. So we have built this enterprise muscle and it is trying to build a little bit of a flywheel of sustainability.

And to your point on Conmigo, Conmigo has a very real marginal cost to it. It costs on the order of, let's call it, $10 to $15 per month. And so all of these school districts that have come to us, we are talking to them of like, hey, we need to at least cover our computation costs.

And then we do expect that the underlying computation costs are going to come down. And so we are having these conversations on our team right now of like, okay, right now it costs, let's call it $10 per user per month, just for the computation. We have to charge that for Conmigo, not for BaseCon Academy. But let's say that cost goes down to a dollar per user per month. Do we just lower it to a dollar or do we lower it to $2?

And then we use that incremental dollar to fund our R&D. I mean, I look at this, the potential here, right, and what you've developed and the amount of money you've put into it already. But I wonder whether there's a world where you're. take a product like Conmigo, which, if it's as revolutionary as it looks like it could be, I mean, it can really change the face of education. Is there a world where that's spun out into a for-profit business? There's always been this debate, not at Khan Academy, but about Khan Academy, which it always confused folks why we're not-for-profit. You know, we scale, we're tech-heavy.

I live in the middle of Silicon Valley. Most of my friends are entrepreneurs or VCs of some kind. And there's a Harvard Business School case about this, should Khan Academy be for profit or nonprofit. And to me, the arguments for for profit have historically been access to talent and access to capital. The reasons to be nonprofit are you really can truly make your mission the bottom line. And I think there's a trust aspect of it as well. You know that our bottom line really is this. It's not trying to improve our EPS or have an IPO. What we have found is we are getting not to just good talent, but to the best talent. People are here for the mission and they're here to work with other people like that.

And so the only reason that I could ever see trying to do some type of spin-out or a wholly owned sub that is for some reason, if we're not getting the risk capital from the foundations or we can't create a flywheel of sustainability, then maybe certain pieces of it. But I'm very afraid to do that. I mean, definitely wouldn't do that with the mothership because I do see how a lot of good comes, I'm a diehard capitalist. My old job was at a hedge fund. A lot of good comes out of market incentives. But I think education and healthcare in particular, market incentives don't always align with our values. So one of the incredible triumphs of Khan Academy has been the millions

of kids who have benefited from it. But in the United States, a huge concern, as you know, and I know you've looked at this, has been the gap between wealthier kids, kids, kids, who are not as wealthy or who are impoverished, a gap in outcomes between black students and other students. How can something like this change that in your view? I mean, I don't want to be too much of a techno-optimist here, but I do want to be optimistic. I mean, and I'm sure you are. How do you think they can kind of change that equation? Yeah.

Some of the stats would shock folks. A majority of minority majority schools. So that's a mouthful. but a majority of minority-majority schools don't offer courses like algebra two, physics, et cetera, not to even mention things like AP courses or IB courses. And so if you're a young African-American student, you don't come from a lot of money,

you go to one of these schools, you might be the next Albert Einstein, you might be the next Marie Curie, but there's no way that you're going to be able to tap in your potential if your school does not offer algebra two. or many times in that algebra two class, because kids are coming from, in many cases, tough circumstances

have gaps in their education before that, the teacher kind of tries to teach to the middle or even to the bottom quartile, saying that those are the most severe cases. And once again, that kid who could have been Marie Curie is not going to be able to prepare themselves, and then when they go to college, they're going to feel inadequate compared to kids

who had much better preparation. So that's always the view of Khan Academy, which is we want to raise the floor. And then we want to provide as many supports as we can. What we view ConMigo is kind of in between traditional Khan Academy and Schoolhouse. World, but that's just another layer of support. Now, in order to access all of this, you still need Internet access,

you still need a device. And that's why we work so closely with school districts. You know, the technological side in the U.S. has gotten a lot better over the last 10 years. But that's why we're working with school districts to try to get more kids at that level of engagement. Sal, just a broader question about AI, as these products, as other AI platforms become better and better, you can imagine a future not too far in the distance where things like writing won't actually be that useful. Like, you know, we think of this as a skill. We think of great writers as people who've honed their craft, who have talent. But, I mean, if these large, you know, language models can absorb everything ever written in human history, well, you can also imagine that that skill and talent won't be particularly useful in a short period of time.

Do you agree with that? That was my initial thought slash fear slash concern. The more that I thought about it, and I'm writing a book about all of this right now. So I've been interviewing folks. And, you know, I interviewed Kevin Ruse, who famously, you know, New York Times writer, who had that face. famous conversation with Bing's Sydney and it was trying to convince him that he didn't love his wife and he loved her and all of these, all of these things. And I asked him that same question. And I thought he brought up a very good point. I mean, he's literally a writer. And he's like,

look, to be a great writer or journalist, there's the writing part of it and then there's the journalism piece of it. And sure, AI could help you with the first draft. It can help give you feedback on your writing. But it's going to take a lot to get to great writing. It might not get quite there and that whole piece of journalism, like talking to the right people, being creative about how you get your information, connecting the dots, AI is not going to do anytime soon. And I think every job has that aspect of it. So I actually think the imperative isn't that, oh, kids aren't going to learn to write. I'm worried about the kids that only learn to write okay. Well, to some degree, they might be empowered by something like AI because now their writing is not

going to hold them back in other domains. But anyone in the, let's call it the writing lane, they're going to have to become better. They're going to have to move into the editorial role to be able to know how to manage to create great writing. How do you know that the AI has done a good job and that you can't coach it to be even better? You won't make someone an editor of a newspaper unless they could be one of the best writers themselves. And I think the people who really leverage AI well are going to be the people who can get into that. How do you manage the AI, how do you put the pieces together? And how do you do this things that AI won't be able to do? So anyone who has a child, who knows a child that's been in school during the pandemic,

knows how much education has evolved, right? I mean, many kids are now issued laptops, and not just in private schools, but also in public schools. And so much of the work is done through Google Classroom, right? And parents can interface with it. And it's just a completely different world. Forget about what they're learning. It's how they're learning. But I suspect that what you're talking about with things like Conmigo and other AI platforms, what we're about to see is something that is like going from, I don't know, like the Stone Age to the Industrial Revolution. Is that a fair analogy? I think you might be right. You know, there's a world now where you can talk to historical characters. And we're probably three to five years away where you could share a room with Benjamin Franklin or

Julius Caesar or Cleopatra, and immerse yourself. It'll be literally the holodeck from Star Trek. I didn't think that was going to happen in my lifetime. I now think that's going to happen in the next five years. I think for the most part, this is positive, because the students who were really motivated, you know, there's a class of students when they were in their math class, and I was in this class of students.

I imagined how I was describing the universe. And I was like, oh, this is so beautiful. This is so elegant. Wow, this is so connected. And that motivated me to go get through some of the grunts. and the really hairy equations and all that. But for a lot of kids, they didn't see the beauty in it,

and so they weren't motivated. But if you're motivated, you can power through anything. Same thing on history. When you just read the history book, like, ah, this is dull. But if you're like, wow, this actually happened to real people, and I actually know about it. And could you imagine what it would have been like to be in that moment when Caesar crossed the Rubicon,

you know, what was going through his mind, then all of a sudden history comes alive and you become incredibly motivated. I think AI will really help unblock folks, as a tutor, but also really help motivate. I mean, imagine having a tutor that not only can do what Aristotle did, but it actually can be Aristotle sometimes. It can roleplay with you.

It can take on tone that really captivates you. This all would have been science fiction a year ago, and it's literally happening now or in the coming months. It puts to rest the question, if you could have dinner with five living or historic people, who would they be? Because you can actually do it at some point. You can actually do it. You know, having the real dinner will always be better because we, you know, obviously the AIs are, they're interpolating or extrapolating from, from.

So there's definitely some imperfections. And but yeah, you're right. In terms of the richness of experience, it's even better than the dinner because probably in five years you could put some goggles on and have the dinner actually in Rome in the first century BCE as opposed to on your dinner table. It's amazing. It's absolutely amazing. Sal Khan, thank you so much. Thanks for having me, Guy. That's Sal Khan, founder and CEO of Khan Academy.

Hey, thanks so much for listening to How I Built This Lab. Please make sure to follow the show wherever you listen on any podcast app. Usually there's just a follow button right at the top so you don't miss any new episodes, and it is entirely free. If you want to contact our team, our email address is hivt at ID.wondery.com. This episode was produced by Alex Chung with editing by John Isabella. Our audio engineer was Neil Rauch. Our music was composed by Rumtin Arablewe.

Our production team at How I Built This includes Chris Messini, Carla Estevez, Casey Herman, J.C. Howard, Liz Metzger, Sam Paulson, Carrie Thompson, and Elaine Coates. Neva Grant is our supervising editor. Beth Donovan is our executive producer. I'm Guy Raz, and you've been listening to How I Built This.