When the Khan World School launched last year as a partnership between the Khan Academy and Arizona State University Prep Digital, it promised a new way forward for virtual schooling. One year later, Sal Khan and Amy McGrath rejoin me to share how the first year went—including some stunning first-year results, lessons learned, and what the next year ahead will look like as they enroll students for the next school year. As always, subscribers can listen to our conversation as a podcast, watch it below, or read the transcript.
Michael Horn: Welcome to the Future of Education, where we are devoted to helping create a world in which all individuals can build their passions and fulfill their human potential. And to help us further that idea, today we have two guests who are returning guests to the conversation: None other than Amy McGrath, the COO of ASU Prep Digital, and Sal Khan, of course, of Khan Academy fame, and the two have partnered together, as many may recall, on creating the Khan World School. And first, Amy, Sal, great to see both of you. You're almost a year in, I think, to a full school year, if you will, of the Khan World School. How's it gone? How's that first year been? Congratulations.
Sal Khan: Well, and first of all, Michael, you had a hand in this, for those listening. I don't know if it was a year and a half ago, two years ago, we were looking to... Obviously I've been thinking about how do you put all the pieces together to really reimagine learning for a long time. We had started Khan Lab School out here in California many years ago, but we said, hey, after the pandemic, there's an opportunity to do online schooling, but do it right, do it so it's not just students by themselves in isolation, nothing but screen time, but actually get the best of both worlds. And I was talking to a bunch of folks of just who's done interesting things, and Michael, you said, you should talk to Amy McGrath. And Amy and I had a conversation, and by, I don't know, minute 45, there was a little bit of like, hey, you wouldn't happen to want to do this together? And Amy, I'm paraphrasing, was like, I was hoping you would ask.
Amy McGrath: Yes, indeed.
Sal Khan: And we ended up... So Michael, you are definitely the matchmaker here. But I've got to say that it has, I don't want to jinx it, but it has gone surprisingly well. I give a lot of credit to Amy and Betsy and their team at ASU Prep and ASU generally, where I and the Khan Lab School team came to them with a three to five page vision document, which really talked about personalization mastery, things that I've been talking about for many, many years. You could leverage tools like Khan Academy for students to learn at their own time and pace. We leverage another sister nonprofit that we started called Schoolhouse.world, where students can not only get free tutoring, but even better, prove their knowledge, which actually a large number of universities are now taking very, very seriously.
But we want to anchor it not just an asynchronous activity, but seminars where you're actually connecting with people. But when you're on Zoom together, you're talking to each other, you're digging into the issues of the day. Borrow a little bit of the Oxford model where students are having small group tutorial sessions, they have check-ins with their advisors. And things like the humanities, where Khan Academy has some content, but not all the content, having reading lists, and borrow a little bit from some of the great books type programs in curricula, but also giving students choice on how they do that. And Amy and Betsy and their team have executed.
I think, incredibly, we had 50 9th graders in this first year, and I tend to run pretty optimistic and these are students from all over the world as we talked about. I mean, Amy can give the exact numbers, but it's a good chunk from Arizona where it's free, where it's leveraging the charter of ASU Prep, but there're students from India, there're students from all over the country, there're students from all over the world. And the data in terms of their growth has been almost hard to believe, so I've been very careful on how... But we're looking at two to three times growth in math, we're looking at three plus times expected growth in reading. Language, arts, don't even want to say it because it's so large, five times expected growth. So these are all... Even if you discount those numbers... I mean these are the real numbers, I'm not making these up. Even if you discount them by 50%, they would still be hard to believe numbers.
But I think the other side of the coin is equally interesting, is not only are these students performing at a very high level of growth regardless of where they started, but also they seem happy. They seem part of a community. It really is defying a lot of the stereotypes folks have about online schooling. They really feel like they're part of a rich community. I've been doing a monthly seminar with them, and I got to say, every time I talk to these students, I have hope for the future and I definitely have a lot of conviction on this model.
Horn: Amy, what are your reflections on one year in talking about those... Those results are incredible. I had not heard those before, by the way. Sometimes I'm suspicious of the growth rates relative to expected value, but that's also because you never hear anything anywhere close to what you just said. Amy, what are your reflections a year in as you close out the school year?
McGrath: There's so much that we've learned, but certainly both qualitatively, and obviously as Sal just mentioned, quantitatively. It's all about the outcomes for the students, and they came in voracious and high aptitude learners and still demonstrated that much growth, so we are so proud of the students. But they are enjoying themselves. We've had a lot of student focus groups and parent focus groups and seminar is where it's at. These kids so enjoy even the access to Sal on a monthly basis, to talk just about big world problems and healthy discourse. It's one thing for us as adults to try to myth-bust around the perception of online learning as being detached, and then to have kids who are sending us videos about why they feel so close to their community, to their houses, and how they can't wait for more students to join. There's eight countries represented. There's really genuine and authentic relationships that have formed in this initial year, and so it truly has been... I don't think we could have even written the narrative better than the students are reporting back, so we're very proud.
Horn: I want to come back to the word, surprisingly well, Sal, that you just used, but I'll let you say why you chose that word in a moment. But you just mentioned something, Amy, about the houses, and I actually think this is one of the, that and the seminar, are some of the coolest pieces of what you all have put together. This notion that... I think there's four houses or something like that that you've constructed? So what is that experience like? What is this community that these people across eight countries get to build?
McGrath: We really believe in this idea of interdependence on each other, so in order to really engage you need your peers to engage. You can't be doing it on your own. And so even though 50, really, is a small community of learners, we then shrink that even more and have smaller houses where they get to know each other on a more personal basis, do projects, teach each other what mastery is like, work on peer mastery, and then they also have the opportunity to get together in the Socratic seminar sessions that have all 50 of the students. We want there to be multiple opportunities to engage in smaller and more intimate settings as well as in the larger, but the students are... It's been fun in the focus groups to hear them say, "well, I still get to stay with my house next year even though new students are joining, right?"
They have this sense of pride and they've really enjoyed getting to know each other. And it's also fun to watch them play off of each other as the year has progressed because, it's certainly a progressive autonomy type of situation, but in the beginning they're sort of watching each other's vibe and, is this okay to speak back and how do I say this question in the right way? And now there's just more of a general comfort, but a resounding theme is respect for one another and for different opinions and the desire to grow and expand their thinking. That has been a real powerful piece in terms of a learning of Socratic seminar.
Horn: Sal, you get to teach them once a month you just said. Maybe not just in that experience, but more broadly, what has been the student experience from your perspective? How do they describe it? What have they gotten to do that maybe they're also surprised about? Not just you saying, wow, it's gone surprisingly well.
Khan: Yeah, I think the things that... The student experience, this was all an idea on paper, and what we've really tried to do is go on this journey where there's Khan Academy that's being used by thousands, tens of thousands, hundreds of thousands of classrooms around the world, and there we have all of these efficacy studies that if students are able to put in even 30 to 60 minutes a week it leads to really big gains. Then in 2014, we started Khan Lab School, which at the time was a very small lab school out here in Northern California, that's where my kids go to school. And we started saying, well, what if you really leaned in on personalization and mastery and having one-to-one connections with advisors and community? Everything Amy just mentioned. And we saw even more dramatic results. Khan Lab School pretty consistently sees one and a half to two and a half grade levels per year, especially in math. We need to, frankly, at the Khan Lab School side, get better at measuring what's happening in other places.
But when we wrote the vision document for Khan World School, we decided to go even closer to vision, which is, we kind of took out the traditional schedule. We did anchor it with these synchronous seminars, my seminars are not the only one, they don't happen just once a month, that's just when I show up. They're happening regularly several times a week. But if we anchor with that, have other synchronous moments where you can connect with your advisors, your tutorial groups, but the rest of it is very much up to the student. This was a bit of an experiment to see if it threads the needle in the right way. And what we're seeing over the past years, I think we've hit the sweet spot, where the average student is spending about two hours of synchronous time with their community, but that two hours, I would argue, is more connected than the five or six hours that a lot of students might spend in a traditional classroom. Because in a traditional classroom, oftentimes you're somewhat passive, sometimes you're lost, sometimes you're bored, you're not necessarily engaged, and you'd be lucky if you have even 30 minutes to an hour of where you actually are engaged in a live classroom.
Here, they're having two hours of pretty small group, very engaging, it could be with their tutorial, which is very small group, meeting with their advisor, it could be the seminar, which is also reasonably a small group, where they are engaged, where they're talking, they're discussing, they are exchanging ideas, working on things together. So I think having that anchor, but then having the rest of your day where you can engage asynchronously, you can message each other, you can do ad hoc meetings similar to the way we would do in a workplace if you want to meet with someone, you want to discuss something, and it really feels like that's the right amount of flexibility. And we are now... It's feeling right, and as I mentioned the numbers in terms of performance are, even for me, almost hard to believe. But because of that, we're actually actively trying to take a lot of what Khan World School is doing and bringing it back to KLS. KLS was, to some degree, the inspiration of Khan World School, but Khan World School has gone further and now we're trying to bring that back into the physical setting as well.
Horn: That's pretty cool actually, to hear about how they inform each other. Iterations over there then can feed back and then vice versa. It's almost like siblings competing with each other or twins, Amy, maybe in your case. But the question I'm curious about is, what have you learned about why students chose to enroll in Khan World School, and what's the message for those who are thinking about what they want to do for their schooling next year that are thinking, is Khan World School right for me?
McGrath: I can answer, at least start us. We imagined a school of students that wanted to go really fast, that had a ceiling in traditional settings, that if we eliminated some of the walls and the ceilings then they could put together their own learning experience that's all anchored to mastery. And that is the case, but not every student wanted to go as fast, and some of the students really needed a lot of training around what does it look like to leverage mastery and have few rules in how you demonstrate mastery. Students are coming from really structured environments and realized, I don't know what pace I'm on. I need more. I need more. And so this year, we are going to be working in the first couple weeks with our newest students on what it looks like to be involved in this model. That was a key learning for sure.
The students wanted to sprint, but also were afraid to unleash because they might not get an A, they're going to get a high mark of high proficiency, et cetera. And so some of the things that they were used to being incentivizing to them moving forward were different, so that was an interesting piece.
I think we're also seeing that some students really want to focus heavy on one subject a week at a time and then do the next at another week. And so some of their styles that have come out have been key, and not just us assuming this is what sprinting looks like, but asking the students, how do you pace your week? Where do you want to learn? And truly, our learning guides have been key architects into this experience and they're really helping to inform, I think, some of the Khan Lab School. Very excited that the collaboration is so much that we're going to have a pod in Mountain View, California, with Sal's son as a part of this, as a Khan World School pod at Khan Lab School. And so you can imagine that environment and that cycle of being able to make one another better and draw from each other's assets is going to be pretty fantastic.
Khan: Yeah, I'll just add... I mean, you could tell I'm voting my beliefs by convincing my kids. And look, my son, who's going to be a freshman next year, you asked me what type of student I think... The number one reason we yell at him in the house, my wife's like, Imran, stop coding. He's that kid. Maybe it's not too much of a surprise for folks. He likes to solve puzzles, he likes to code, he likes... But he's curious about things. His favorite thing to do with me is we'll go on a walk, he's like, "hey, is it okay if I have some questions, some science questions?" I'm like, yeah, it's okay. Let's talk about science while we're walking around the park.
But I would say that a student who is feeling under stimulated, who wants more flexibility, who wants to dig into their passions, who, as Amy said, who doesn't want a ceiling, who wants to be around adults who don't discourage asking questions but encourage asking questions, even hard questions, even questions that adults might not know but the adults are saying, "well, let's figure that out together. Let's see if there is a way to do that." But at the same time, who wants to become part of a community and wants to be around other similar students.
I won't say this is an outright school for high performers. I think all kids who come to this school could end up, in fact we're seeing that in the results, they can end up being high performers, but we really want to... But there are some high performing students who really want to be told every moment of the day what they should do, and it's probably not good for them, but for students who feel like they can do a lot more, who want to have more of a sense of purpose, who want to have a sense of being able to engage with the world and not be limited in any way, I think this is a good fit.
Horn: It's interesting hearing you say that, Sal, because I think that was part of my wonderings as well, is this for the Montessori students, say, who wants to continue that freedom? But it's actually also, maybe yes, but also for the student that's feeling boxed in by the experiences that they've had and sort of looking for that escape valve. Amy, you just said something really interesting though, which is that there's going to be a pod in Mountain View on the site of the Khan Lab School for Khan World School. So this gets us to where I want to go, which is what's next year going to look like? It sounds like those 50 9th graders are going to become 10th graders, there's going to be a whole new entering class and there's going to be some variation in the experience. Tell us more about the pod. What else is coming down the pipeline for next year?
McGrath: A lot to be excited about. We are expanding. We really wanted to start small and be obsessive about the quality and the opportunity for our students, and it went so well that we are going to expand our offering to 6th through 12th grade. We expect about 50 students per grade, so 350 total. The idea to pilot a Khan World School with a physical presence in Mountain View is just that, to pilot. What does this look like from a scheduling perspective and a hybrid learning environment? When are students going to come to school? When are they going to stay back? So I think that's part of our future, is honing it and mastering it in this next year in a safe and collaborative environment, such as Khan Lab School, and then offering this more broadly, nationally, to progressive school leaders that have the right space and have the right structures to allow us to collaborate.
One thing that we're really careful about is not diluting the experience and handing it over to someone else. We're somewhat desiring to be controlling of the model so that it is a purist mastery model and so that... One of the things that we keep talking about is these students, they're not necessarily coming in as crazy high aptitude learners, they're just willing. They know how to self-regulate, they are curious and they desire to be there. And so there are so many of those students that are untapped that may need a physical place to report. And so as we desire to scale this to thousands and tens of thousands, we're trying to create models in various modalities that are going to be able to cater to those various needs.
Khan: I'll just add to that. I'll keep a heaping praise on Amy and Betsy and the learning guides and the whole team there because I... and Michael, you know this, we've both written a lot about mastery learning, which is really just the idea that students have the opportunity incentive, whatever their assessment level is, to improve it, to get to a level of proficiency or mastery. It usually goes hand in hand with things like competency based learning, which is really, hey, it doesn't matter that you took algebra, what matters is do you know algebra. And obviously you could have multiple shots at goal on this. It doesn't matter if it took you a year to learn it, two years to learn it, or two weeks to learn it. But it's one thing to talk about personalization and mastery and all of these things and making things student-centered, it's a whole other thing to implement it.
And what we've seen even at KLS, Kahn Lab School, is that there's a lot of gravity towards going back to the traditional that tends to be a little bit more, the adults direct what a student does on a day-to-day basis. The assessment, maybe it doesn't feel as mastery as we would like, it's not as personalized as like... So to be able to get such a, I would call it a pure but well-thought-out implementation, with such great results is pretty darn amazing. And so that is one of the reasons why I'm so excited about scaling it behind, bringing some of that back to KLS, where we can create that program and see how you can do a hybrid between a physical layer to supplement the Khan World School experience, which I actually think is going to make it a lot easier for a lot of folks.
A lot of folks around the world are looking to do innovative schools, but it's hard to come up with the curricula and the grading and the college admissions and all this stuff. Now they can just take that from KWS, we imagine, and then focus on the physical setting and the environment that they want to create.
Another dimension that I think is going to be really exciting, as you know, Khan Academy has jumped in with both feet on artificial intelligence, both Khan Lab School and Khan World School were the first places where we were able to try things out. It's had a huge impact on what we're actually creating. And I think as we go into next year, you should never use technology for technology's sake. You should think about what is the problem you're trying to solve. But there's some very clear problems that we've been trying to solve at Khan Academy, KLS, and now KWS, that we think artificial intelligence could play a very, very, really interesting role.
One, it obviously can act as a tutor for every student that has even more of value and a personalized mastery framework. It has powerful capability to act as a teaching assistant, help create lessons on the fly, make them dynamic, help prepare students in ways that would've seemed science fiction. I just gave a TED Talk and I shared a story of a young girl from India who goes to Khan World School, her name's Sanvi and she's reading The Great Gatsby, and she was trying to figure out why he keeps staring at the green light. And she says, "oh, I realize I could actually talk to Jay Gatsby via Khanmigo." Via our AI that we have on Khan Academy, and then she has this conversation with Jay Gatsby where he is like, oh, old sport. Yes, what do you want to talk about? And she's like, "well, why do you look at the green light?" He's said, that's at the end of Daisy Buchanan's dock, and it represents the distance between where I am and where I want... Anyway. And then she apologized for taking up the simulation, Mr. Gatsby's, time.
But that's the type of thing that... We're rolling it out to schools around the world, especially around the country. But I think Khan World School is going to be the epicenter of not just throwing in technology in there just to look modern, but to do it in ways that are really thoughtful where we can start to do even more mastery based learning and things like writing. And on Khan Academy, it's pretty straightforward to do personalized mastery learning because it grades it for you and there's a very deep item bank so that you don't see the same thing repeats. But with essays, Khan World School has already been doing it, but if you had an essay and you're not quite mastered yet, in theory, you should be able to edit it and then someone should review it again and keep assessing it. But how do you make sure those assessments are consistent and transparent and you get as many shots at goal as needed? That's very resource intensive, and we have been doing it at Khan World School, but maybe artificial intelligence has a role here.
Artificial intelligence can do higher order tasks, even in math, that we couldn't traditionally do, where a student has to explain their reasoning, a student has to give a proof, a student has to design an experiment. And once again, the artificial intelligence can both work with the student, but also assess the student and also communicate with the adults on what the adults can do to better support the students.
Horn: It's cool to hear you say all that, and I'll tell this story on air: One of my daughters was home sick from school a week or so ago, maybe two weeks ago at this point, and she was asking me questions, and not quite pestering, but about various things. And I said, you know there's this talk of my friend Sal that I need to watch of him giving this TED Talk, which I knew I should watch, not just the explanations you had forwarded me originally around Khanmigo. And so we watched it together and then she said, "okay, my turn." And so she hopped on, we had the invite to Khanmigo, and by the end she was full on programming by the end of the day. It was an incredibly productive day and she said, "see, dad, I can be productive and learn outside of school too." And I said, "of course you can. That's awesome." It was incredible to watch the interaction between the two.
And I think the other thing that you just pointed to is, you've almost built this dual engine, seems to me, Sal, of Schoolhouse.world on the one hand of peer-to-peer almost assessment, but peers who are denoted as masters of the things that they're assessing, coupled with perhaps the artificial intelligence. The amount of information we're going to be able to get about people's learning and true mastery thresholds in a much more seamless sort of authentic way seems like it could really dramatically change the learning models and sort of our mindsets around what teaching and learning even looks like. I'm sort of curious, how far has this played out in your head about what's possible?
Khan: Yeah, and I know for folks listening, it could get confusing because we have an alphabet soup of organizations that we have, with Khan Academy at the core, and now we have the schools, Khan Lab School, and what we've been talking about, Khan World School, and Schoolhouse.world. During the pandemic, it's always been my dream... And we've always seen it in classrooms, that the best classrooms don't just have students on Khan Academy, but the students are able to engage with each other and provide that extra support towards each other so that you don't just have one teacher in the room, you have 30 teachers in the room. You can get to a one-to-one that way. And we said, well, what if we could scale that broadly, especially during the pandemic where folks were at home and they needed to feel more connected.
So we started a sister nonprofit, Schoolhouse.world, its mission is connect the world through learning. And the whole idea is peer-to-peer tutoring. And we said, well, how do we certify whether someone can tutor, whether they know their material and they can tutor? Well we said, what if they took Khan Academy assessments that come from a very deep item bank, so you could take it as many times as you want and not see repeats, and record their face and their screen while they explain the reasoning out loud. And then that video artifact, if they get to 90%, if they don't, they can try again tomorrow with a different test, but if they get to 90%, it gets peer reviewed, to your point, Michael. And if it all looks good, and obviously Khan Academy is doing the grading, the peers are making sure that the person didn't cheat, they followed the protocol, then we say, okay, you know unit one of calculus, you can start your journey to tutoring. Then there's training and all of that.
But at the same time, during the pandemic, universities like University of Chicago, MIT, reached out and said, hey, could we use this as a way for college admissions? Because one, we love this way that you're assessing folks. You're not just making them fill out a Scantron, you're making them explain their reasoning, and they also have to review other people's work. That by itself is awesome. But then these days, especially at some of these elite schools, kids with maxed out grades and standardized scores are almost a dime a dozen. We want to see the kids who can teach this material, who can go off to other people and explain it and invest the time. And to your point, that has created this really powerful... It's not the same scale as Khan Academy, Khan Academy is tens of millions of folks, this is tens of thousands of folks, but it is scaling quite fast.
But both Khan World School and KLS have anchored deeply on this for our math, because one, it's a way to show that you really know it, but then we also want our students to go on that journey to eventually tutor other students, which by the way, there's now 16 universities including MIT, UChicago, Yale, Cornell, who have become part of this consortium who are taking these types of signals seriously. But to your point with artificial intelligence, we're just starting to explore how the artificial intelligence can run a simulation with a student, how it can ask them questions. You can imagine pairing that maybe with what we're doing at Schoolhouse, where you record yourself engaging with artificial intelligence, maybe in the humanities, it can run essentially like a thesis defense with you, and then maybe that gets peer reviewed by humans. And so we do see something interesting developing, to your point.
Traditional Khan Academy with the exercises and the practice, we think that's going to be relevant for a while because it is the most curated and clear indication, but then you have a layer of artificial intelligence, both on the support side and on the potentially of the assessment side eventually, and then you have the human support and assessment at the highest level with Schoolhouse.
McGrath: I'd love to just comment on that as well. As a learning designer, watching students have access to every one of the tools in real time that Sal just described, it's been phenomenal to watch 13, 14, and 15 year olds with Khanmigo, and the respect that platform has for the student on their learning journey is pretty phenomenal. I mean, there's so much controversy around this right now, and it's challenging me as an educator. I mean, the whole theory that the brain who does the work, does the learning. I truly do still believe that, but I think we're seeing it in a different way now.
I think some of the future of all these tools that Sal is just talking about, this universe of assets, so you've got AI sitting between the student and the teacher, and it accelerates exponentially their ability to not just be waiting on them and on the adult, on the human. It's making connections and pushing the student further and we're really seeing that actually happen. And so the students are still doing the work, but they're actually doing a better work product and going back to it, asking the bot how to deepen and expand their knowledge base, which is, I think, improving the ability to master as well. So very, very impressed as an educator in architecting new models that this is a good thing for students.
Horn: It strikes me that that's where the focus is as we start to wrap up here, Amy, which is it's not just taking the tool and just putting it blindly in an existing system, it's building new models around it. And that's what fundamentally you all are demonstrating here, is that Khan World School and Khan Lab School are these Petri dishes in which to reimagine what schooling looks like. And then with this technology, the scale that you can start to do this at, to allow communities all over the world to form around this, it gets pretty exciting pretty fast. Take us through where you're thinking... You're enrolling 6th through 12th grade coming up, you just said, where does this go from here in both of your minds as we wrap up?
Khan: Well, I view Khan World School as a Formula One car analogy where automotive companies, they have the Formula One cars to show what's possible. If you take the best of all the technology, the best engineering, the best drivers, what you can do, that's what Khan World School is. What's interesting is it could also potentially scale. It's not going to scale to millions likely, but definitely tens of thousands over time. But I think... And it also creates a very powerful outlet for students today who want to be challenged. And I think these are going to be the kids who start the Googles of the future, who cure diseases, write the great novels of the future, and even if we can reach thousands of them, I think that's going to have huge impact on society. But I think even more importantly, it sets an example of what's possible in the broader world.
The timing of Khan World School existing hasn't been better because of all of the artificial intelligence work that's happening. We need to show the world how this can be used for good, how it can accelerate outcomes and be very clear-eyed that we're not just alluding ourselves, that we're measuring it and we're seeing that it's actually working. There's a very interesting moment in EdTech that's happening where classic EdTech has always been, hey, you've been running a school a certain way, here's a tool that might improve things and here's our efficacy studies, et cetera. What's happened with AI, it's been like a big bomb in the education system where it kind of broke education or it is breaking it where people are like, what do we do with term papers? What do we do in a world where kids can cheat and ask ChatGPT things? And we think we have a solution here where not only could we mitigate any of those risks and make it safe, but we can dramatically accelerate and do things that would've seemed utopian, honestly, even a year ago.
McGrath: Yeah, so increasing opportunity is what Arizona State University is all about. And so this K-12 concept of ASU prep, creating new models, and now being able to partner with someone as prolific as Sal to design a model that is giving students hybrid flexible choice and allowing them to sprint is so exciting. We also are doubling down on early college. Truly eliminating that line and bringing college to 13, 14, and 15 year olds to master and to continue to just... No longer is it all breath. There's a lot of depth. And when a student knows what they want and can just go after it and tutor themselves using AI, show mastery with their peers on Schoolhouse.world, have access to state-of-the-art university content anytime, anywhere, it's pretty phenomenal. So I think it's the best of what we know as education experts all in one menu.
The desire and the goal is certainly to expand that and to get more students on board. So 350 next year, I don't know Sal, what do we say? A thousand the following year? That's the pace that we would love to go because we think this is so great, and as the outcomes are speaking for themselves, this is something that we now... We hypothesized it was going to go really well and that we knew how to design a powerful model, now we are seeing in real time that it's good for kids, so we're excited to get it into more hands.
Horn: Khan World School one year later. People should keep checking it out and you all keep expanding so that you bring these pods to Massachusetts so my girls have some good options for high school. I'm excited about where this is going. Sal, Amy, keep up the great work and thanks joining me.
Khan World School, One Year Later