Duolingo's CEO on how he built a $700 million company and why he gave away his first invention to Yahoo for free
Chantal Heijnen for Duolingo
- Duolingo cofounder and CEO Luis von Ahn is a crowdsourcing pioneer and the recipient of a MacArthur Fellowship, commonly known as a "Genius Grant."
- Von Ahn grew up comfortably in Guatemala City but traveled to the United States to join the American tech scene.
- He is one of the creators of "captcha" technology, as well as its "recaptcha" successor, which helped digitize a year of New York Times articles in a week.
- His language learning app, Duolingo, has millions of users worldwide and is preparing to go public.
- He shared his lessons on humble leadership and why he's trying to convince people to move from San Francisco to Pittsburgh.
You may never have heard of Luis von Ahn.
But if you've ever verified that you're not a robot online, you've probably used something he created.
And that's fine by him. He doesn't read stories about himself, and when he says he's not motivated by money, he sounds sincere.
"Everything that I've done has been with the goal of just having a lot of impact on a lot of people's lives," Von Ahn said on Business Insider's podcast, "Success! How I Did It."
Von Ahn is the CEO of Duolingo, the popular language-learning app now valued at $700 million. He's also one of the guys who developed "captcha," the online prompt that asks you to type in a word or series of letters to show you're not a robot.
Von Ahn grew up in Guatemala and came to the United States to be part of the tech scene. He was driven by a goal of getting millions of people to work together online. It made him a pioneer of the now ubiquitous crowdsourcing movement, which is where we started our conversation.
Listen to the full episode here:
- Audible founder and CEO Don Katz
- Media mogul Tina Brown
- Planned Parenthood president Cecile Richards
- Vimeo CEO Anjali Sud
The following transcript has been edited for clarity.
Luis von Ahn: Early on it was trying to get millions of people to work on something together - basically crowdsourcing. And now it's trying to get millions of people to learn something with Duolingo trying to get millions of people to learn languages.
Rich Feloni: So this idea of getting people to work together on the internet, is that something that still drives you even if it's in other projects?
Von Ahn: Yes. It's something that I think about all the time. This is something that I worked on a lot in the early 2000s: How do we get millions of people to work on a common goal?
Falling in love with computers in Guatemala City
Feloni: So this love of starting companies and working in tech, did it begin as a kid?
Von Ahn: I think so. I mean, technology definitely started as a kid. When I was 8 years old my mother bought me a computer. I wanted a Nintendo and she bought me a computer instead and I was pretty pissed off. I was forced to learn how to use it because I wanted to play games with it. She got me a Commodore 64 and I learned how to use it. So that's when I started really getting into computers. The starting companies thing, that didn't really happen until I was in my 20s.
Feloni: What was it like growing up in Guatemala City?
Von Ahn: I had a pretty normal childhood when I compare it with other people here in the United States. I was fortunate that I was kind of upper-middle class. Both my parents were medical doctors. Guatemala is a very poor, third-world country, and when I was growing up it turns out that there was a civil war going on, but I never really experienced it. It's this interesting thing. People sometimes are, like, "Wait, didn't you grow up during the Guatemalan civil war?" Which is true, I did. But I didn't experience it. I had a pretty normal childhood. I think I was very fortunate to be shielded from all of that.
Feloni: Do you remember as a kid when that computer, that your mom got you at 8, when it went from a lousy gift to something that you actually got really excited about?
Feloni: So it was cool to be the guy everyone went to for these.
Von Ahn: Yes, and people would show up at my house, and usually these are much older people. It's not like they were 40 years old, but, you know, I was 10 and these people were, like, 17. And my mother just said, "I don't really want to have these teenagers coming into my house randomly." So at some point I had to do all my dealings in the front door. I wouldn't let them in. It's like selling drugs.
Feloni: That's funny. So at this point, this experience of being this 10-year-old video-game bootlegger, did that end up inspiring you, like, "Hey, maybe I could when I grow up, actually make a career out of doing stuff with the computer."
Von Ahn: Yeah, you know, I thought when I get older I could probably do something with computers.
Becoming a crowdsourcing pioneer
Feloni: So you end up at Carnegie Mellon University studying computer science, and then you get into the early days of crowdsourcing. Could you tell us what captcha is?
Von Ahn: Sure. So this is something that happened in my first semester of the Ph.D. program at Carnegie Mellon University. The guy who was the chief scientist of Yahoo came to give a talk at Carnegie Mellon about, like, 10 problems that we don't know how to solve at Yahoo. I listened to the talk, and when I went home I tried to work on all of these problems.
I couldn't come up with any great ideas for any of them, except for one. And the problem was they had these people who would sign up for free email accounts by the millions. So somebody would write a program and they would use this to send spam, because what happened is that each Yahoo account only allowed you to send, like, 500 messages a day. And if you're a spammer, and you want it to send 10,000,000 messages, you would get a couple of million email accounts and from each one you would only send 10 messages or whatever. These people wrote programs to obtain millions of email accounts, and Yahoo did not know how to stop them.
I started telling my Ph.D. adviser at the same time about this problem, and it occurred to the two of us together that the thing to do was to have something that distinguishes whether the thing creating an account is a human or a computer program, because computer programs can obtain 2 million email accounts easily. But a human, after obtaining 30 email accounts, they get bored and they stop. And so that's where we came up with this thing called the "captcha," this test that can distinguish humans from computers.
Feloni: When Yahoo approached you with this, did you think, "OK, I'm going to make a deal here, maybe get rich"?
Von Ahn: No, this did not happen, actually. We were just so happy that we had made this thing that Yahoo was ready to use. We gave it to them for free. To this day I hadn't really much thought about making much money. I mean, even with the game bootlegging, it was not a moneymaking thing. I was just happy to have a bunch of games. It was similar here, and neither my Ph.D. adviser nor I thought of making too much money here, or make any money. We gave it to them for free and we were just happy that they were using it.
Feloni: So you don't have any regrets over that?
Von Ahn: No, I don't. I mean, it ended up working out pretty well for me, so I don't have regrets about that. I'm happy with it. I mean, I would've done things differently if it was now, because I now know how the world works a lot better, but at the time, I mean, it's not like I sit here regretting it.
Feloni: Are you not naturally business-oriented? I mean, you're saying that you didn't want to sell things as a kid and it didn't even cross your mind to sell to Yahoo.
Von Ahn: Yes, I don't think I am naturally business-oriented, and this is something, I mean, even with this, my latest venture, Duolingo, it took us a while to start actually making money. By now we are making significant amounts of money, but this has not been my driving force. As my driving force has really just been making products that had a lot of people use. And yeah, I don't think I am naturally money-oriented.
Feloni: When did you realize that your driving force was just making technology that lots of people use?
Von Ahn: After helping create captcha, the satisfaction that it gave me to see that, every day millions of people would use this, and that everybody I talked to had seen these or had come across them - or, well, most people actually hated them - but that just gave me a lot of satisfaction. That a lot of people had used this. And from then on, everything that I've done has been with the goal of just having a lot of impact on a lot of people's lives.
Feloni: Did you ever get annoyed at one of your own pieces of technology, having to answer one of those captchas, that maybe it's hard to understand?
Von Ahn: For sure, and then the idea was just copied a lot and many of them were not actually developed by me. I mean, sometimes the letters are very highly distorted. The most annoying thing for me is always just buying tickets on Ticketmaster, because they give you, like, 90 seconds to complete the transaction, and there you are trying to enter your credit card and everything, and then you can't read the damn thing. So yes, I became annoyed at that a few times.
Feloni: In 2006, you got a lot of attention when you got one of those MacArthur Fellowships, which people often call one of those "genius" grants. What was it like when you got that phone call?
Von Ahn: Yes, that was a completely unexpected. I mean, I was just in my office one day and I got this phone call and the guy just said, "Have you ever heard of the MacArthur Fellowship?" I said, "Sure, I have heard of it." And then he said, "Well, I'm happy to tell you that you've been selected for one of them." And it was completely unexpected. They're very secretive about how they choose their recipients. And yes, I was obviously very proud of myself, but every time that I get one of these awards, I also feel guilty. I don't know why. There's probably years of therapy there to fix that, but I also kind of felt guilty. I don't know why, but that happened.
Feloni: What do you mean by that? Guilt over what?
Von Ahn: I don't even know. I just feel like I don't deserve it. Like, "Why am I getting it and not somebody else?" I just get this feeling, it goes away, but I do get this feeling whenever I get like an award or the first time I saw myself when I originally made captcha, pretty soon after there was a pretty large spread on The New York Times, a big story about this and there's a big picture of me there. And the first time I saw myself, the feeling that I got was, "I don't deserve to be there. I don't know why I'm there." It's like guilt.
Feloni: After that, when you started getting more attention, did you find a way to make peace with that somehow?
Von Ahn: No, I've never really found the great way to do that. I don't read stories about myself. I just - that's kind of how I do it.
Feloni: So you just end up having to use it as another vehicle for spreading the technology, and if you have to talk you have to talk.
Von Ahn: Yeah, I mean, for sure and it is a good way to spread things. In the case of Duolingo, for example, a lot of the spread, we've never done any paid advertising or anything like that. So a lot of the way, which Duolingo spread is by having, for example, news stories about it. So it is very valuable. It's just not something that I particularly love.
Feloni: When you had the captcha technology, you were saying that you realize that maybe there's something else that you could do with this. What inspired you to make this "recaptcha" technology?
Von Ahn: Yeah, that was this second go at the captcha. Above five years had passed, after I had helped invent the original captcha, by then essentially every website used these. And I did a little back-of-the-envelope calculation that about 200 million times a day, somebody types one of these on the internet. And at first I was pretty proud of myself, because I thought, "There's a lot of people every day that are using something that I did." But it was the case that most everybody that I talked to hated doing these. And I also started feeling guilty that each one of them took about 10 seconds of time. And if you multiply 10 seconds by 200 million, you get that humanity as a whole was wasting like 500 thousand hours every day doing these annoying captchas. That's when I thought, "Is there a way in which we can do something else with this technology?"
And that's when it occurred to me that it could have a second purpose. And that's where the recaptcha project was born. Where the idea was that, as people were typing these distorted characters, the idea was to also get them to help us digitize books. And the way that worked is, at the time, there were a few projects that we're trying to digitize all of the world's books. Like Google had one, for example, the Internet Archive had another one, where the idea was taking all the books that had ever been written and scanning them to put them on the internet. Now, the way this works is, you take a book, you scan it, and now scanning literally what it is, is it's taking a digital photograph of every page of the book. It gives you an image for every page as an image with words in it. The next step is that the computer needs to be able to decipher all of these words. But for older books, where the ink has faded, the computer cannot recognize many of the words.
So the idea was to take all of the words that the computer could not recognize in the book-digitization process and to get people to read them for us while they type recaptchas on the internet. So basically, whenever you were typing a recaptcha, these words were coming directly from books that the computer could not recognize, and we were using what people were entering to help digitize this, and that was the idea of the recaptcha project. It became very successful. And at the height of it we were probably digitizing about a 100 million words a day, which is the equivalent of 2 million books a year. We started out actually by helping to digitize The New York Times. Old editions of The New York Times were being digitized by people on the internet typing captchas.
Feloni: How did that conversation with The Times come about?
Von Ahn: That was actually complete luck. I was giving a talk somewhere in Texas, and it's a very large audience, about captcha, and what we could do with it. And then it turned out that the CTO of The New York Times at the time was actually in the audience, and he came to me at the end, and he said, "I know what you can digitize with this. We have this huge archive of all editions of The New York Times." It was like 130 years or something like that. And pretty soon we had struck a deal on a digitizing all their content, but that was kind of luck in that respect that the guy was sitting in the audience. And then after that, midway through that project, actually Google decided to buy recaptcha to help with their book-digitization efforts. So, that's how that went. So Google now owns recaptcha.
Feloni: And with that, The Times case, you had some great results, right?
Von Ahn: Yeah. I mean, this is the beauty of captcha. It really is very accurate. It was taking us about a week of time to digitize an entire year of New York Times content.
Feloni: Oh wow.
Von Ahn: Yeah, and over time, the rate of digitalization was becoming faster and faster because more and more websites were using this version of captcha as opposed to the previous ones. I know that the rate became a lot faster when Facebook signed up. So Facebook started using recaptcha as opposed to their own version of captcha. And as soon as they signed up, basically every person who signed up to Facebook would help us digitize one word of either a book or a newspaper, and that's a lot of people.
Feloni: And with recaptcha, you sell it to Google and you actually made a good deal out of this. What had happened in terms of your perspective on your own approach to technology that changed where you actually were going to start selling some of your creations?
Von Ahn: I don't think anything had changed. At some point, I sort of understood more how the world works - and it's not like I'm against making money. It's just never been a driving factor for me. And so this was a pretty good deal for both sides. I mean, Google really needed this technology to help digitize their books, and for us it made perfect sense because we had the New York Times contract, but it wasn't like we had access to millions of scanned books like Google did. So on our end, we just had one piece of the whole process, and Google had all the other pieces. So it made a lot of sense.
Changing language learning with Duolingo
Feloni: When did you realize that you wanted your next project to be Duolingo?
Von Ahn: It was soon after having sold recaptcha to Google. I thought to myself, "What am I going to do now?" I had to spend a couple of years at Google after the sale, but at some point I just realized I really like to be my own boss. I really like doing stuff by myself. So I just thought, "OK, I'm going to start a new project. What have I always really wanted to do? What has been my passion?" And I just always thought that education was my passion. I've always loved learning stuff, so I thought I wanted to do something related to education, but then I started thinking: "OK, education. Let's see."
First, I came back to my upbringing in Guatemala. It's a very poor country, and I saw firsthand how those who have access to education can basically do very well in life. And in countries like Guatemala, a lot of people talk about education as something that removes inequality or that brings the social classes together, but I always saw it as the opposite, as something that actually makes an inequality, because this was the case for me. I was able to get a very good education, and, because of that, you're able to continue having a lot of money. Whereas people who don't have very much money in countries, particularly developing countries, barely learned how to read and write. And because of that they can never kind of get out of poverty. I just got to the point where I wanted to do something that would give equal access to education to everybody. That was my driving factor when I wanted to start Duolingo.
And then the next step is education is very general. And so I thought maybe I'll just do something a lot more specific. I tried to just narrow down to one thing, and I thought of learning languages, which is huge everywhere in the world except for the United States. Learning languages in the United States is not just as big. In most countries learning English is a huge thing. In developing countries where English is not the commonly spoken language, those who know English usually can earn up to twice as much as somebody who doesn't know English. It was a thing that would really try to help or had the potential to help a lot of people get out of poverty. That's the driving factor for creating Duolingo.
Feloni: How many languages did you speak as a kid?
Von Ahn: As a kid, just English and Spanish. My native language is Spanish, but I learned English very early on, so I just spoke English and Spanish. Then I tried learning French when I was in high school, but I failed. My motivations for learning French were not great. I had a crush on a girl who was in the French class, and so I decided to enroll, but I never learned anything. I also failed at getting the girls to go out with me!
Feloni: You met President Obama at the White House a couple of years ago, to talk about Duolingo. How did that happen?
Von Ahn: Yeah, the White House had a few startups come, and it was like a startup show-and-tell. I don't know how they selected them, but we actually weren't told that we were going to go present to Obama. We were just told it's going to be a thing in the White House. But when we showed up, we were all told, "OK, there's a very special guest that's coming. The whole time it's been the case that he was going to come and listen to it, but we just didn't want to tell you guys." We had the opportunity to talk about Duolingo for a while with him, which was pretty cool.
Feloni: What did he say to you?
Von Ahn: Well, Duolingo was mainly an app on iPhone and Android devices. He said he was not allowed to use a smartphone. I guess things have changed - presidents are now allowed to use Twitter. He said he wasn't allowed to use one of those. And so he said as soon as he finished his presidency he was going to try to learn some languages. He was pretty keen on the fact that we were working on education. We talked a lot about how Duolingo is used in public schools. Duolingo is used in about 20% of the US public schools, in language classrooms. That was something else that we talked about just generally technology to improve education.
Feloni: Can this app help replace or supplement existing language courses?
Von Ahn: I wouldn't say replace. The idea is to improve it. I mean, our idea with it is you can learn a language with Duolingo by itself. If there's a teacher available, usually it's better. The teacher-plus-Duolingo is better than either Duolingo or with the teacher separately. For those people who don't have access to a classroom or a teacher, they can use Duolingo and it'll work. But if they do have access to a classroom and a teacher, it'll just work better. That's our idea. There's just some things that with an app you just can't quite do, that you can do with a teacher. Teachers are really good at answering some types of questions. Those are really good at motivating people. Classrooms are extremely huge motivators in part because you're being forced to just show up. They're extremely huge motivators.
Feloni: Are you going to be taking Duolingo public?
Von Ahn: That's the goal. We're not quite there yet, but my best guess is we're two to three years away. Our revenue is growing quite significantly year on year. That's the goal. It's one thing for it to be the goal and another thing for all the stars to align, but that's the goal.
Feloni: Are you worried at all that you might not be able to maintain a social component of it if you have to keep checking on quarterly earnings?
Von Ahn: Yes, that is a legitimate worry. So far we've been able to navigate that pretty well. It's of course very different. I cannot tell you that I understand the public market super well since I've never been the CEO of a public company, but so far this has worked pretty well with private investors. We have a lot of investment from venture-capital firms. Google is one of our investors, for example. It's worked out pretty well in terms of being able to do both, having a social good and making revenue. But yes, it's a worry. It's not our biggest worry, but it is a worry.
Feloni: Would you want to remain CEO if it becomes a public company?
Von Ahn: Yes, I think so. I think you're also supposed to say yes to that question no matter what, but yes.
What inspires him
Feloni: I saw that, at one point, Bill Gates tried poaching you. What happened there?
John Lamparski/Getty Images
John Lamparski/Getty Images
Feloni: What do you guys talk about when you meet up?
Von Ahn: Of late it's been really Duolingo. The last time was in January. He was interested in learning more about how we're using artificial intelligence in Duolingo, to try to teach better. He was lamenting the fact that a lot of education companies particularly are claiming that they're using AI in very sophisticated ways, but in reality it's pretty unsophisticated. He was trying to understand all the places where we use AI. I would say for Duolingo, we're about medium-sophisticated. There's no company that's using artificial intelligence to teach that is, like, "Oh my God, this is completely groundbreaking in terms of AI, and we're going to substitute all human teachers, et cetera." I don't think we're anywhere near that yet.
Feloni: Why have you stayed in Pittsburgh?
Von Ahn: Good question. I like it, and it's really on the up. At the beginning, I was in Pittsburgh because of Carnegie Mellon, which is just an amazing computer-science program, and we started it here. We didn't see any reason to move at the time, and we thought that being near Carnegie Mellon would allow us to hire a lot of great engineers, and it's been true. And over time it's just that Duolingo grew bigger and bigger, and at some point it just became impossible to move. We may start another office somewhere else, but moving at this point, it's just too many people. That's kind of why, but it's worked out pretty well. Being in a city that's on the upswing is pretty good because you're just getting a bunch of highly talented people wanting to move here. That has helped us.
Feloni: Yes, so is there an advantage to being in Pittsburgh that you might not get maybe in the Bay Area?
Von Ahn: There are ups and downs. Some of the pros: being near Carnegie Mellon really helps. We're able to hire really amazing people from there. Another big advantage is just cost of living. We're getting just a good number of people who are just either moving from the Bay Area or just don't want to go there to begin with, fresh out of college, because here you can get an apartment that's pretty nice and you can just pay, I don't know, a thousand bucks a month, and that is completely unheard of in the Bay Area. Many of our employees - I'm going to assume maybe half of the employees in Duolingo own a home, and that's completely unheard of in Silicon Valley. Those are the advantages. We actually just put up a billboard on US 101 near San Francisco that just says, "Own a home, work in tech, move to Pittsburgh." Then it just says "duolingo.com/jobs."
I don't know what that billboard will do. We hear that enough, that we thought it was worth putting up the billboard.
Feloni: We talked about some plans for Duolingo maybe in the next few years or so, but what you have coming up this year?
Von Ahn: This year we're spending a lot of time teaching more advanced stuff. Duolingo has been really good at beginner to intermediate learners. We're going to be concentrating a lot on more advanced learners. We're also spending a lot of time on just making it a strong business, our monetization. One thing that I'm very excited about is one of the ways in which we monetize Duolingo is we have ads at the end of the lesson. The ads that you're going to be getting are actually going to be video ads in the language that you're learning.
For example, if you're learning Spanish, you may get a short video of a Coca-Cola ad in Spanish. I'm pretty excited by that because our users love these. People actually want to watch them and pay a lot of attention. This is one of those things that is good for the user, but also good for the advertiser because these are not ads that people want to skip. They actually want to figure out what is said in the ad.
We also just launched a podcast, which did really well. Season one ended, but while we were still in season, it was in the top 10 podcasts the whole time, now it's ranked No. 40 or something of all podcasts, which is pretty good given that it's a Spanish-learning podcast.
Feloni: I'm actually taking a look at an episode of your podcast right now, and it said that in one of the episodes that before you left Guatemala that your aunt was kidnapped. Is that a true story?
Von Ahn: That is a true story. Our podcast, the way it works, they're all real stories, and on the last episode of the first season was actually a story about me that happened to me, as the year right before I left for college here in the US, my aunt was actually kidnapped in Guatemala, and so it's the story about how that happened.
Feloni: If you chose this as the story to tell, is that something that impacted you heavily?
Von Ahn: Yeah, it did impact me. It was actually one of the decisions for me to leave for the United States. There are other reasons why I wanted to leave, but this was kind of the last nail in the coffin for me deciding to leave Guatemala. I was like, "Oh, my God, now people are getting kidnapped here too?" She was kidnapped, she was gone for about 10 days. This was a money thing. In different countries, kidnappings are for different reasons.
Feloni: Was she returned?
Von Ahn: Yes, she was returned. This was a business. Kidnapping was a business, it is not good business to kill your victims. She was taken; she was actually taken care of pretty well. They got all her medicines and they fed her pretty well and then they just waited until the ransom was paid and then they gave her back.
If it's not interesting, he won't do it
Feloni: Has there been a moment in your career where there was a particular insight that really ended up guiding you?
Von Ahn: Can you use curse words in this podcast?
Feloni: Yes, go for it.
Von Ahn: Yes, I learned this. This is a good, pithy one that always stuck with me. It's about hiring people. It is really detrimental to the organization when you hire people who are not very nice, but when you're hiring, and particularly for a startup, you're usually pretty strapped for time. You're just, like, "Oh, man, we've been looking for this position, whatever head of marketing or whatever it is. We've been looking for this position for months and this person's almost good but they just have this one little problem." There's always a huge desire to try to hire them and maybe that problem will not manifest itself in the company, et cetera.
What I have learned is this pithy line: "It's better to have a hole than an a--hole." That has worked pretty well for me. Duolingo is a very nice company; everybody here is pretty nice. We're very proud of the fact that we have no a--holes.
Feloni: How do you define success?
Von Ahn: Oh, boy. For me, I think it's being able to do what you love. There's the thing that you love and then there's your job. To me, success is kind of when you're really able to do what you love.
Feloni: Has that understanding of what it means to be successful changed over your career?
Von Ahn: Yes, it has. When I was younger I thought, "Well, OK, success probably means you drive a Ferrari and you have a big mansion somewhere or something like that." That has changed, and it's been changed by just looking at a lot of very "successful people" on what types of things they appreciate. I've been around enough very successful people, and the things that bring them joy are just not - it gets tiring and boring to drive around in a super-fancy car or whatever. That's what has changed it.
Feloni: So it's you also finding ways to just stay interested?
Von Ahn: Yes, although for me it's been pretty natural. Maybe this is a bad thing for a CEO to say, but I just don't do the things I'm not very interested in. I have the luxury of that when I'm not super interested, I just don't do it. The fact that I have that luxury, to me, is real success. It's just I don't do things I'm not very interested in.
Feloni: Thanks so much, Luis.
Von Ahn: Thank you.
- Apple iPhone 13’s A15 Bionic won’t be based on 3nm process, confirms TSMC
- The utter chaos and confusion over Remdesivir in India is making the COVID-19 second wave worse
- Atlassian, the maker of Jira and Trello, is looking to hire 300 engineers in India this year
- Night curfew hours extended in UP's Lucknow, Prayagraj, Varanasi, Kanpur City, Gautam Buddha Nagar, Ghaziabad, Meerut, Gorakhpur, Shravasti and Moradabad districts
- Around 1,700 people test positive for COVID-19 in Hardiwar Kumbh Mela from April 10-14