Salman Khan is the man behind Khan Academy, a 2009 Tech Award winning site with 12+ million views and 1200+ 10-minute "videos on YouTube covering everything from basic arithmetic and algebra to differential equations, physics, chemistry, biology and finance". We talked with him about building "the world's free virtual school", the potential of open-access learning, using the format for sustainability debates, and challenges in growing a non-profit from zero to global impact.
Hassan Masum: Salman Khan, thanks so much for joining us today. First of all, tell us what Khan Academy is, and why it is significant.
Salman Khan: Well, its current incarnation is twelve hundred videos on YouTube - and other places, but YouTube is its main source of people connecting with it and finding out about it.
And there is a whole software piece that generates problems for people, and starts them off with one plus one equals two. The software piece will take them up to about Algebra II, but the video piece will start them at the most basic one plus one equals two level and take them well into college level calculus, linear algebra, differential equations, physics, chemistry, and biology.
The goal is to create the world’s free virtual school. Right now it's more content, but there is a kind of community forming around it. And between the community that is forming around it, the video content, and the software, we can do data manipulation and understand how we can keep tweaking it and making it a better platform for people to learn on. I envision it as the world’s free virtual school.
HM: Interesting. Do you have any metrics for its success so far?
SK: In terms of raw numbers...if you go to any YouTube channel you can see total video downloads. Some of the university open courseware efforts are older than the Khan Academy...if you just look at the raw numbers, only one of them is a little bit higher: MIT had 12 point something million cumulative views as of today, while Khan Academy had just over 12 million. All the other university open course efforts, whether it's Stanford or Berkeley or Yale, they're all a good bit lower than that.
What I have started to do is keep track of the daily change, and on the daily change Khan Academy is getting about 50% more views per day than even MIT which is considered the standard-bearer for open courseware. So the first metric is that I think I can legitimately now say that it is the most widely used open course video content, despite being one person's effort as opposed to a major research institution's.
I think what these universities are doing is tremendous, but Khan Academy has got a lot of traction that way. It has as I mentioned reached 12 million video views, it has about 80 or 90 thousand unique students a month using the site, and that's growing by 15 to 20% per month.
So that's just in terms of the raw numbers. For me personally, I think the bigger impact is the testimonials that I get, and if you go to the site you will see a lot of them.
HM: It is rather amazing that this site which has been around for not too long, and is more or less the effort of one person so far, has achieved this success. One really wonders what the "secret sauce" is. In other words, why is the site so successful, and what do you personally think makes your videos so enticing to watch?
SK: I am always trying to answer that question for myself, because I don't ever want to lose the secret sauce! (laughter) I think it’s a combination. First, it's the length of the videos: the fact that they're ten minutes makes them very digestible. They are almost like the Dan Brown Da Vinci Code type novels, where you feel like you can almost read another chapter and then go to bed.
And I think it’s this notion that you don’t have to sit through a 90 minute lecture to get to exactly what your weak point is. You don't have to sit through a 90 minute lecture on calculus if you just don’t understand implicit differentiation. The way a lot of the other content is categorized, you don't even know which lecture to watch to really hone in on your weak point.
The format, even more than the time limit, is probably what people find engaging. It’s very informal - they can tell it’s just me, probably at my house, doing these. They started off for my cousins...it feels like someone's older brother or family friend who is with them tutoring.
I started off with no fancy equipment. You don’t see my face - you just hear my voice. You have a black background, and I use bright colors to get some nice contrast to make it visually interesting. I've had a lot of feedback from users that hearing the voice with the black background kind of feels like it’s happening in your head. It feels like the voice in your head. And it’s very conversational - it’s not your traditional slightly cold or stilted lecture that you might normally get.
I got one note a couple of days ago saying that “you teach like a friend, not like a teacher". I think that lowers the stress level when you learn mathematics. You can compare it to the offline medium - the notion of being able to pause and repeat, and watch whatever you want.
I think that the other element is its breadth. For a lot of older learners that are going back to school, their main intimidation is jumping into a class and being completely lost. But when you go to a site and see "okay, there are some videos on calculus and whatnot, but the videos go all the way back to one plus one equals two" - you know that there is a point you can jump in. Or even if you jump in at a higher point, and get a little bit confused with exponents, there is a video that can help fill in the gap.
So I think it's also the breadth of the content. People feel very good about going there, and knowing that there will be something that they can connect with.
HM: So granularity, format, and breadth. It seems like those three elements have the common feature that they all make the energy barrier to getting involved lower.
SK: Yes, absolutely! I think the format, the tone...if I had to guess one, it would be the tone of the lectures.
Actually, I’ll add one more, not to throw too many out there: the focus on intuition and retention and connections, as opposed to the mechanical memorization side that many students often find themselves doing. I get letters time and time again where they're almost angry - they say "our teachers think that they need to teach us this at a superficial level so it will be easier, but it actually becomes a lot easier when you learn why you are doing it, or why this formula actually makes sense intuitively".
And so I think people appreciate the idea that they are getting the real deal. They are getting a deep understanding, as opposed to something superficial.
HM: Absolutely! So, not to get you into trouble, but I wonder if you could speculate on why it is that those elements are in relatively short supply in formal education.
In other words, if you think about Khan Academy, it's one person - and you were actually doing this part time for quite a while, though now you're full time. And you've created this resource which, as you said, is now getting the same number of hits as some of the world's best-known institutions.
On the other hand, you have this huge formal apparatus with many dedicated people in it - probably tens of billions if not hundreds of billions of dollars per year of investment, and there hasn't really been this kind of resource created. What if any lesson does that hold for future educational efforts?
SK: The simple answer is to put stuff out there and iterate, and not have a bureaucratic team that are better at shooting down each other's ideas and constraining teachers. I understand the need to constrain teachers, because you want to have quality control and make sure everyone is being reached. But the negative side is that you're also constraining very good teachers, and you're taking a lot of the humanity out of the lesson.
This happens at the textbook level as well, and the state standards. I think to some degree there are so many cooks in the kitchen that the final product that the student gets is extremely diluted. There's something to be said for fewer cooks in the kitchen - and if they're good cooks, the food will be a lot more fun to eat. (laughter)
That's my best answer. Several states apparently have had efforts along the same lines. The idea isn't mind-blowing: get your best teachers in the state, or in the country, and put a camera in the room - I don't use a camera, but you could put a camera in the room, or use a format like me - and have them teach. And put those videos online, and make them free for the world.
The expense is almost ridiculously low to do something like that. But time and time again, some of these states have contacted me and said "well, you know, it's getting stuck in meetings..." - and they really haven't produced any videos.
The best way to think about it is that it becomes very corporate. There is this view that it has to be very polished, and have computer graphics, and that the teacher has to have a script so that they don't say "um" or make any mistakes. And I think what that does is it takes all of the humanity out of it, and the humanity is what people connect with.
HM: Hmmm, yes. Let's talk about your motivation for a second. You quit a career in finance to do this. Tell us what your motivation was - both looking backward, and also how you think you'll stay the course looking forward.
SK: When I started, the only thing I had to give up was a little bit of my time after work. As you mentioned, I didn't start doing this full time until about four months ago, so I was doing this essentially as a significant part-time activity for about 4 years before that.
My initial motivation was this notion - if you read the website, I started with my cousins and all of that. And that's just tremendously satisfying. When you work at a hedge fund, it's a fun and intellectually stimulating job on a lot of levels, but it's tough to answer that question of "what did you do for the world today?" or "what value did you create for the world, other than for your investors?"
Just in a small way, it made me feel good to start with my cousins. Eventually, when I realized that I could "scale myself up" through YouTube and the software, I'd get notes from random strangers saying how their lives had changed, and how they're going to become an engineer now, and how they loved math for the first time in their lives. When you get those things from people all over the world, I don't think you have to be a Mother Teresa to feel pretty good about it.
When I first started, I thought it was a nice thing to have out there - something that my or your great grandkids could learn from, and it's fun. But over the last 6 months to a year, it started to be pretty obvious that not only is it a nice thing, but it's transformational for a lot of people out there. It's already reaching hundreds of thousands of people, and it could be reaching hundreds of millions.
And it was getting traction. As we discussed earlier in the conversation, it's getting traction above and beyond what major research institutions are being able to accomplish. Even though there is no obvious way in the short term to necessarily give me the type of standard of living that I might have been getting in my career - especially because I was at the point in my career where, if I would have stayed in it, I was at a kind of inflection point there too - it will never provide that type of livelihood. But it's been pretty clear to me that the value is definitely being created. And I just have to figure out some way for society to recognize some small fraction of that value, and at least be able to pay my rent and put food on the table.
I have YouTube ads on, which do revenue sharing at about $2000 a month right now. I'm pretty convinced in the worst case that just that, at the rate the site is growing, would be enough in about a year for me not to have to dip into my savings or my checking account to fund my family and this effort. But I don't want people to have ads for non-educational things while they're trying to learn algebra, or calculus or chemistry or whatever. So I am talking to some foundations, and I'm always open to sources of foundational support. Donations are coming in - I'm actually getting about $1000 a month in donations, and that's growing with the viewership.
So Plan B is donations and advertising - if I were to be really obnoxious about advertising, I could probably be there today, but I just don't want to do that. But if I got foundational support, then I could hire a team of five or ten people, and could really fulfil that idea of a full-service virtual school.
HM: Sounds very promising! Tell us a bit about how, at the personal level, Khan Academy has been perceived. Where I'm coming from is that, often when people go from a relatively, let's say, high-status job as perceived by society at large - doctor, lawyer, engineer, and so forth - to one of these more entrepreneurial efforts...we've heard over and over again how there can be a barrier in getting the message across as to why the effort is important.
And so I'm just imagining "Salman Khan, Hedge Fund Manager", versus "Salman Khan, guy who sits in his basement making YouTube videos..." (laughter) I'm just curious, have you had any challenges at a personal level - or with colleagues, with friends and so forth - in terms of getting the message across as to why this is important?
SK: I think if I had quit three years ago - some of my friends thought I was crazy even when I hadn't quit and I was not showing up at social events to work on YouTube videos instead. But I think by the time that I did quit, especially within close friends and family - a lot of them started using it, or at least they had family friends who started using it and would tell them "hey, do you know this guy? My son is addicted to his videos - could you get his autograph?" (laughter)
So amongst close friends and family, I think they understood it. When I meet new people...my wife is a fellow in rheumatology at Stanford, and when they meet me and say "what do you do?", and I say "well, you know, I teach on YouTube". Then they're like "oh, this guy is freeloading off his wife." (laughter) I never got that when I was working at a hedge fund. Then they were like, "good thing she married him!" (laughter)
I had to get over that, but I'm pretty much over it. And I think the letters I get from students, and getting notoriety in the mainstream press now - that obviously helps with my more superficial friends and family who will only realize the value of things once someone else confirms it.
HM: Fantastic. So in terms of the videos themselves, let's dig in a bit to the topics you've chosen. You were a math / CS major in school, and of course a lot of the videos are on mathematics, physics, biology, and so forth. The thing that those have in common is that they're relatively objective topics, in the sense that there's not a lot of room for different schools of thought on them.
So one question I have for you is, how well do you see this same format working for topics in the social sciences, or possibly topics which are a little more politicized? Do you think the same format will work? If so, why, and if not, what would need to change?
SK: I've hinted that I do plan on doing it, and I think it will work on one level - but I think there's an element where it won't work.
The level where it will work is to give people the baseline. I was talking to someone yesterday about how I would go about giving lectures on the Israeli-Palestinian conflict. That's something where if you just stress one word a little positive or a little negative, you're probably going to anger a lot of people. And I think nothing could be more debatable than an issue like that, or the India Kashmir issue - we could keep listing issues like that.
My take on it is that there is a reality that everyone agrees to. And I can convey that reality, and I can map it out, and I can draw out issues. My job is to be as good as possible at articulating both sides of the story, or three or four sides of the story if they exist. And that's what I think I can do on YouTube in the same format.
I have made a couple of videos like this. One would think that natural selection in evolution is a fairly objective topic, but as you know it's not - it's a highly politicized topic, probably more politicized than some of these conflict regions that we just talked about. I have made videos on that, and I actually made a few videos on intelligent design.
I think one of the problems is that people who believe in intelligent design are just outright dismissed by people in the scientific community, and what I wanted to do was sort of reconcile intelligent design - not in any way to justify it, but to show people who see beauty in things, to say "hey look - I'm not advocating whether I believe in a Creator or not, but if you believe in a Creator, what's better: a Creator who designs the specifics, or a more elegant design that generates the system in which all this complexity arises?"
I try to be very sensitive to all of these issues. The fun thing is that I can put a lecture out, and if it generates a lot of debate - there is a lot of debate often in the intelligent design videos, half the people thought I was a crazy creationist, and half the people thought I was an atheist (laughter) just because of the nature of the topic. But I can keep adding videos. I can say "well, you all said this, and I want to be very clear that this isn't what I said - I said that." And I can keep adding context to it.
I kind of look forward to those videos - I need to experiment with it more, but I think on some level it can be a better format than what you traditionally get. Some of these topics, they're so heated and controversial that I think people completely shy away from teaching them. If you talk to someone about any of these conflicts around the world, 90% of what comes out of either side of the argument is usually just outright factually incorrect. And I think just getting the facts straight would be hugely constructive.
The piece that I think would be difficult is that debate that you have within the classroom, where you're all sitting around a table. But even there, you have message boards now, and people can put response videos. We have yet to see it, but I think it will be a good format for this type of stuff.
HM: It sounds like you're almost aiming for what Wikipedia calls its "NPOV", which is Neutral Point of View - where the goal is not to say that A or B is correct, but rather to portray each side (or all sides) as best as possible.
This seems to also tie in to the possibility of version control of videos. I'm wondering if you see the possibility of having "video sets", or something where you might have a sort of "parent video", and then subvideos on different aspects or different ways of of portraying it.
SK: Yeah, I think that's exactly right. You've seen the site - most of the site's topics are very linear. You start the playlist, and you keep going and it advances. But I think in some of the social sciences, that's going to be the way to go.
You do a high level - this is the 10 minutes of what you've got to know about what happened to the Western World in the 20th century. And then you can do 10 more videos on each of the decades, and then 10 more videos...and keep drilling down into further detail.
HM: Another idea that comes up is the whole idea of 'learning paths'. Typically if you're learning calculus, there's a pretty clear progression of topics. In other areas, it's not so clear what the linear path is.
And yet, at the same time, you can imagine there's a sort of graph of dependencies, where there are topics which logically come first, and topics which come afterward. How do you see putting in advice to learners, as to which videos might logically follow on - you should watch A before B, A after B, or if you like A, then B might also be of interest - how might that be integrated into the site?
SK: The simple answer is 'yes'! (laughter) You mentioned this graph structure, where you think certain concepts might have certain prerequisites, and they feed into others. And you can imagine this kind of node-directed graph map of concepts.
That's actually what the software has. When a user logs in, they can click on something called their 'knowledge map', which is exactly a node-directed graph - the node all the way to the left of the screen is '1+1=2', and if you become proficient in that, it leads to level 2 addition, and level 1 subtraction. You can do it at your own pace, and you can proceed really far down one node while not so far down the other node.
I'm completely with you on that. I want to do the exact same thing on the videos. The modules on the software side have videos associated with them, so it kind of happens. But I want to do that exact same thing, and the direction that I want to take this in is that everything is integrated - so we know what videos they're watching, who they are, what exercises they're doing before and after the videos, and how effective the videos were in getting over the hurdle in that exercise.
And then doing exactly what you said - saying "look, for someone like you, we found that this is probably the next video you should watch". Or "you might want to review this stuff, because we've seen that when a lot of people did so, it led to good outcomes". I'm extremely data-driven, and it's really just a matter of getting down to doing it. But it's completely feasible, completely possible, and I think that's one of the exciting dimensions of something like this.
HM: Let's talk about how this might be applied to, say, teaching environmental or other such topics. So let's assume that somebody wanted to get a citizen's-level view of climate change. The first possibility would be to have simply a playlist: "here's 10 videos which tell you the basics of what you should know."
What comes to mind in terms of how initiatives like Khan Academy might complement what's already out there, in terms of information on topics that are large-scale and long-term, and perhaps where there isn't a clear sense of what the core knowledge is as yet?
SK: This is an area where I'm not an expert, so I'm talking as someone who wants to find 'the true answer'. You hear a lot of noise on both sides. I have an opinion, but I'm not 100% sure. I can go watch CNN and I'll have people screaming at each other - not just CNN, I can watch any news channel and watch people on both sides of the equation equally angry and sounding equally sure of themselves.
And the only fact that I can go on is "oh, this guy's got 80 scientists behind him, and that guy has 10, and the guy with 80 is probably right". But who knows? That's not how we should decide.
I could go listen to Al Gore's talk, which is a little more in-depth. But if it's important for you to prove it to yourself, that still may not meet the hurdle. You'd want to have someone show you the charts, and show you why it's statistically relevant, and why this isn't happening by chance, and why this isn't happening because of the natural cycles of the Earth, and why these claims are statistically implausible. You want an analysis - but you can't get an analysis right now, because it's just in papers and textbooks which, even if you can read them, are going to put you to sleep.
So I think there's a gap there. And I think that's where 10-minute videos would be ideal. I would love the leading climate experts to, in a very conversational style, distill why they are convinced of their argument. And not just resort to "oh, well I've got more people on my side than you do" or "oh no, you're just saying that because it's popular to say that". I want to see him analyze the data and show me why it's statistically unlikely that the other guy is right.
HM: How might this format also be used for topics which are 'how-tos'? Let's say for example that I believe that there's a real problem with climate change, energy use, and so forth. And I want to figure out how to carbon-reduce or energy-reduce my product or lifestyle.
There is a lot of how-to information out there, and I'm wondering if the how-to side is different from the knowledge transfer side in terms of how you would approach the teaching.
SK: I think how-to is on some levels easier. You don't have to necessarily whip out equations or analyze charts - you just show people.
The format I use could be good. You could talk it out and write it down, and maybe it will leave some kind of imprint on peoples' brains. I could also imagine a video that would be equally good or maybe better, where I see "oh, this is how he uses his computer to control the power sockets in his house".
I think this format could be one of many formats. I don't think it's necessarily the format that has to be done for how-to type topics.
HM: You mentioned the developing world as being one key target, and also the possibility of offline snapshots that could be transferred over there. First question on that front: do you have a sense of what kinds of views you've had so far from around the world?
SK: Looking at YouTube data, I would say that probably a fourth to a fifth of views are coming from the non-Western world - you know, outside of North America and Europe. If I had to guess where those are coming from, the ones that are watching us online live, it's probably some relatively affluent people who have both the computer and broadband access. Right now it's not reaching the poorest of the poor. And non-English - it's people with some reasonable proficiency in English.
There are groups - one is called uconnect.org, and the other worldpossible.org - that have taken offline snapshots of the Khan Academy. They're distributing them to villages in Uganda, Ethiopia, Nepal, and Latin America, with the view that it is a 'school in a box' - a very limited one, it's really just content right now.
What I hope to see, and I'm talking with some foundations about this notion, is this idea of - I think broadband access will become pretty ubiquitous over the next 5 to 10 years. Maybe one of the reasons it's not more ubiquitous is that people don't see the rationale for giving broadband access in a lot of the places - "if I give them broadband access and one of the new hundred-dollar laptops, what are they really going to do with it?" There are stories of people using these laptops for lighting, because they don't have lights at home.
What I envision is, if I'm a foundation, NGO, or government, I'd go in and instead of giving every child a computer, I'd set up a village center and have a handful of computers, and hire a local guy or gal to man the computers and make sure nothing goes down. It could even have a slow internet connection. And what you do is, you put the software, the content, locally on their machines. So students in the village - which could be the young kids who you normally associate with being school age, but it could also be their parents - on their own time, they can learn asynchronously.
The problem with traditional schools is, one, you can't find teachers for many of the advanced subjects. Even if you can, it's hard for the students to even go to school, because they have to work - oftentimes in the year they just can't make it. If you have something like this, they all learn at their own pace, and you don't have the teacher shortage issue because you have essentially a virtual teacher.
The big sell for the philanthropists is that, for the first time, they can actually have data on how those assets are being utilized. Right now, if I go build a bunch of schools in India or Africa, I have no idea what's happening with them, especially with the computers. In fact, even in the U.S., in some schools they have no idea how the computers are being used, or whether it's advancing the kids' understanding of anything. But now if you have this 'school in a box' with the software and the content, the data can connect with the mother ship, and the person who's allocating the resources can see in real-time: "gee, these 10 kindergarteners in Malaysia are doing great - let's throw in some more money at that NGO, because they seem to be a really effective one."
The sell really is asynchronous learning, 'school in a box', more economic than the current model, and data-driven so for the first time people know what's actually happening on the ground.
HM: There are interesting questions about to what degree we might want to have topics that are more directed at development. So for example things like personal finance, or basic agricultural knowledge, or basic health.
Similarly we might want to find the equivalent of Salman Khan in each of these countries, who might him or herself be the one who makes it even more personalized to that region. What are the key next steps or challenges on that front?
SK: Well, regarding the internationalization of it, there are volunteers who have come out of the woodwork who started translating. There's actually a Khan Academy in Espanol, and it's a chemistry teacher from Uruguay. It comes off quite well, far better than I would have expected - it sounds like he's teaching it in Spanish, it doesn't sound like a translation.
So there's potential for that. I think it would be even better if there were people native in the language who maybe even redid the videos. It may be that in certain cultures a different approach might be better than the way I approach things. But at minimum, they should be able to be translated by volunteers, and you really don't need that many volunteers to do it.
That's one big step that seems to be happening, although my experience right now is...volunteers are very well-intentioned and I will never turn away well-intentioned inspiration and energy, but it's often hard to get full completion and follow-through. If we get foundational support or some governmental support, we could institutionalize that, and make sure it happens.
The other next step is, as I mentioned before, to hire a small team of 3,4,5 developers. Then we can really support the more interactive parts of the virtual school, well beyond just the videos.
HM: You mentioned contributors. What would you most be looking for from contributors, aside from competence and follow-through?
SK: Actually, that's pretty much it! (laughter)
I've got a lot of friends who I think would be brilliant teachers, on the software side brilliant people to help with that - they want to, but you don't get the follow-through. I don't know what the secret sauce is, but I think on a volunteer level, follow-through is huge - you can't underestimate that. It's sharp people who have follow-through, who really connect with the vision, because that's what's going to create the follow-through.
It's difficult for me right now, because every day I get emails from quite a few people, and to manage that volunteer process would take me away from the videos and doing what I need to do. I'm still struggling with that - on what is the best way to utilize this resource out there, that people are giving out of their own good will. I'm open to suggestions.
HM: It seems as though one ready avenue might simply be to try to find those excellent teachers who are already paid for pedagogical activities. If you found the world's best astronomy teacher, could he or she be the one who might take on the videos for that?
On the follow-through side, that might help to alleviate some problems. I'm thinking of groups like the Teaching Company, where they have videos on a lot of topics, mostly for adult learners. So that kind of tactic, where there's a single person who comes forward or even who you approach, who might be a subject matter expert but also a fantastic teacher - someone who really gets this approach and runs with it.
SK: Yes, I'm very open to that. I've been talking with some people who clearly have the expertise - Ph.D.'s in X, Y, or Z where it would take me years to get that expertise, if I even wanted to go there. And they volunteered to be that person. I said, "you should make the videos no matter what - even if it has no association with the Khan Academy, it will have value, and people will watch them."
If they do have the same 'feel' as the Khan Academy videos, then I'd be more than happy to include them. And 'feel' doesn't mean good or bad - I'm not making value judgments. There is a kind of branding here, in that if there is someone other than me teaching, I want students to feel like they're getting kind of the same thing - the same conversational style, something that's not too pedantic but at the same time is not sacrificing rigor.
I'd make sure that people know that this is from this other teacher out there, and they'd get the credit for it. The purpose of having the Khan Academy involved is so that it gets proper exposure.
HM: You're actually then mitigating the risk of them not doing it - you just ask them for the playlist, as opposed to good intentions.
HM: Last few questions...I'm curious about challenges and lessons, and what you see looking forward.
So first question, Salman, is what do you see as being your key challenges over the next couple of years?
SK: I think the number one challenge is making the finances sustainable, as we talked about already. I don't think it's an insurmountable challenge - I think it's going to happen - but obviously once that shoe falls, it's in a sense downhill from there. I quit my job because I feel like it is already downhill, and has reached that critical awareness where it's picking up on its own. But to get some funding on a significant level would be a huge boon to the effort.
When I started this, it was amazing - a lot of people would come to me who were skeptical. "How are you going to convince school systems to do this? Why is this different? Why do you think you're going to be able to do something that all these other people have tried and failed at?" Regarding the school systems, I'm just doing it out of that context, and I think that was the right approach. As soon as you try to interface with the bureaucracy, the bureaucracy will take over your effort, and your ability to produce will go to zero.
As far as the question of differentiation goes, I think that's been answered, and gotten confirmation from usage, the media, and even other institutions - showing that this is something real. The big thing is just making sure it's sustainable, and hopefully that won't be too hard a challenge.
HM: And what lessons have you learned doing this?
SK: You know, I've done other entrepreneurial things in the past, and this is by far the most pleasant of all the entrepreneurial activities I've ever done. And the reason is a couple of things.
One is, in the back of my mind I always had very ambitious or even grandiose ideas, but I was happy with small successes. I was happy with just reaching my cousins, and just being of use for my cousins, and then just being of use to the 20 or 30 people who were watching it on YouTube. So I think when you're happy with the smaller achievements, you're not going to go through this emotional roller-coaster where you try to start the next Google - it's either a billion or nothing.
So that's one pretty important lesson I learned. Second is this idea that if you do something that you find generally rewarding, you've got to stick with it. Don't necessarily give up your day job. There were moments three or four years ago, where I got really excited about Khan Academy, because I got a little bit of a slap on the back from somebody. I was thinking, "this is going to be huge - I'm going to quit my job today!" And it was important for me to go home, talk to my wife, calm down, and realize that you shouldn't get too optimistic too fast - we need to eat and live. The idea of being able to do both at the same time, and incrementally on a small scale - I think that's also a huge thing.
Probably the single biggest thing that made this pleasant was that I thought very early on, would I rather be a dot com billionaire on the cover of Forbes, where my main success is the number of zeroes in my bank account? Or would I rather make a good living, a respectable living - maybe what a professor might make - but have something out there that's being used by millions of people, and that will be used maybe for many years after I'm dead? And when I put it that way, the second seemed a lot more appealing to me.
Making it a non-profit made it a much more pleasant experience. One, when you approach people, they know where you're coming from. A lot of dot coms or for-profit companies said "oh, we're out to revolutionize education, we care about kids." But you look at their capital structure, and you know that's not the direction their capital structure is going to push them in. With Khan Academy, it's very clear that this guy is not doing this for any other reason. And so people come out of the woodwork to help you.
The other positive is, I'm not paranoid about competition. I embrace competition. I don't even view it as competition - I want other people to do this. If someone else says they want to do videos, I tell them how to do it. I'm coming from a point of view of abundance, as opposed to scarcity, and I don't think anyone is going to undercut my ability to do this. I'm giving it away for free!
The funny thing is that I'm fairly confident now that if I were to get competitive about it, that it would be difficult for someone to turn Khan Academy into the next Friendster, by introducing the next Facebook. Even if there are other Sal Khans out there, and I hope that there will be, who have their own styles, and their own base of students who really appreciate it - I think there will always be some subset of the world that will say "hey, I like how this Sal guy teaches, and I connect with him on some level." It doesn't have to be all - it could be only 5% of all students, but that adds up to be huge.
HM: Some words of wisdom there. Last question and last word to you Salman: tell us what your dream is looking forward.
SK: If you allow me to get as grandiose as I'm capable of getting, I imagine a network of initially one center - I won't call it a school just yet - a center in every major city, and eventually every village in the world, where students go.
When you go to that center, it would look like a fun, engaging, one-room schoolhouse if you will, where kids of all ages are. But they're not just limited to that school - they could interface with the software and the videos, and other members of the online community. Essentially those are all rooms in one worldwide networked campus.
And the software and the videos and the online piece of it - it does all of the analytics, it takes off the plate all of the stuff that the computers can do. And the rest of the kids' time is freed up to do truly creative things, and truly interact with each other, and truly mentor each other.
And it's cross-generational, cross-age. Each center can have one paid director who is a very inspiring mentor, who is there to guide the students in whatever endeavors they take, but Khan Academy is the core infrastructure that connects everything, and frees people to pursue other creative activities.
HM: Salman, thank you so much for speaking with us.
SK: Thank you!
(With thanks to Misha Ishaq for transcribing help, and to Sebastien Paquet and Michael Nielsen for the inspiration behind the interview.)