The Best One Yet - Exclusive 🦉 “Education Super App” — Duolingo CEO Luis von Ahn announces next product on TBOY

Episode Date: April 22, 2025

How do you get someone to learn everyday? Make it addictive… in a good way. Luis von Ahn has mastered engagement with Duolingo, transforming it into the largest education app on earth with over 40M daily users. And in our exclusive interview, Duolingo CEO announces the next product: Chess.In this episode, you’ll hear…Why the CEO (proudly) calls Duolingo “addictive”Why he will never open an office in Silicon ValleyThe psychology behind push notifications, streaks, and the unhinged owlDuolingo’s viral super bowl commercial (aka “The Growing Butt”)Using fire to fire fight in the fight against social media “If it’s not fun, it won’t work”Why “screen-time” isn’t a problem — it’s inevitableWhy artists will beat nerds in the age of AIHow millionaires on Duolingo subsidize migrantsWhy Duolingo employees didn’t want to make money (at first)Watch the extended version on Youtube: https://www.youtube.com/@tboypod —-----------------------------------------------------Subscribe to our new (2nd) show… The Best Idea Yet: Wondery.fm/TheBestIdeaYetLinksEpisodes drop weekly. It’s The Best Idea Yet.GET ON THE POD: Submit a shoutout or fact: https://tboypod.com/shoutouts FOR MORE NICK & JACK: Newsletter: https://tboypod.com/newsletter Connect with Nick: https://www.linkedin.com/in/nicolas-martell/ Connect with Jack: https://www.linkedin.com/in/jack-crivici-kramer/ SOCIALS:Instagram: https://www.instagram.com/tboypod TikTok: https://www.tiktok.com/@tboypodYouTube: https://www.youtube.com/@tboypod Anything else: https://tboypod.com/  Hosted on Acast. See acast.com/privacy for more information.

Well, the biggest thing, I keep telling people and nobody seems to listen to me. First person I ever had to fire, we're adding another subject, and it is not one that many people have expected. I could see his butt crack. And then it started growing, like his butt started growing. Started really realizing, okay, we can be a little more savage. I actually believe that as AI matures, what's going to be most valuable, the only thing he wants in life is for you to do your lesson.

And the obstacle is that you're lazy. This is bonjour, bonjourno, Maharabi, Marhabi, nihow, aloha and Konichiwa. Our guest today can say hello in every one of those languages. True.

Probably Dothraki, too. Because Duolingo is the largest language app on the planet used by 40 million people every day. But Luis von Onn aspires for more than just language learning. Because he wants his app to become the world's first super app

for all things education. Our guest today, Luis Van On, was born in Guatemala. but launched this tech business in Pittsburgh PA. Although his app uses all the growth hacks that were invented in Silicon Valley. He created a mascot so popular, Duolipa is trying to date it. An unhinged owl mascot. That's such a meme. It got its own skit on SNL. And on today's show, Luis will announce Duolingo's next product.

That's right, Duolingo's next new thing will be revealed right here on Teboy. Yet he's today, we are sitting down with the Lord of the Language. The Sultan of Streaks. The Baron of Being Bilingual. Louise Von On is the co-founder and CEO of Duolingo. There ain't any subtitles in today's episode yet. He's, nope, but watch out for that unhinged owl. Because that thing is a raptor with some major boundary issues.

Today's interview with Luis of Duolingo is the best one yet. Louise, thanks for being with us. Thank you for having me. That's quite an intro. Welcome, man. Great to have you. Really great to have you. Although, we have been doing extensive research on you for weeks,

and we did discover you're a people pleaser. Is that true? I am definitely a people pleaser. Yes. Guilty as well. Full disclosure over there, Jack. Have I ever told you this story about why I am, Nick? It's what my dad taught me.

Why you're a people pleasing? Yeah. So my dad told me his favorite book growing up was this book by Gail Sayers, and it's called I Am Third. It basically says, like, prioritize your friends and family before you take care of yourself. And so that's how I was raised. And yes, I'm very concerned with how everyone else is doing. So you're saying it's nurture, it's not nature when it comes to people pleasing.

Yeah, but of course, we all know the pitfalls of being a problem. people pleaser, right, Louise? Louis, how did your people pleasing begin? The origin story. I'm not entirely sure. I'm sure it has something to do with how I was brought up, but I always wanted to please people. Well, now I feel I could be very selfish to just ask things of both of you guys. One of the big problems for me is I've gotten a lot better, but as somebody who runs their own company and I'm a people pleaser, I sometimes have to tell people things they don't want to hear. So I am, I've gotten really good at sugar-coding stuff. Oh, yeah? But that's not good. It's not good. But that is that, you know, I've had to unlearn that.

What's your sugar-coded compliment sandwich, Louise? No, I mean, at this point, I don't do this. But I will tell you, for example, the first person I ever had to fire didn't understand that they were being fired because I sugar-coded it so much that they showed up the next day. Oh, God. They showed up the next day. And I'm like, I thought I fired them. So I've gotten a lot better at being direct. But that has taken a lot of training.

Okay. Well, in the meantime, Louise, I'm in the envious position. of having two guys who are people pleasers on my right and on my left over here. So midway through the interview, I may ask for a sandwich or something. Fifteen years before this song, two boys from the Northeast met in the dorm. They had an idea to cause a cultural storm. It's the best one yet, but the best is a norm.

Jack Nick, that's it. I don't even think they need to practice. 50% that's a fat tip. Tea Boy City on your at list. If you know, you know, because we're rare to go. We can't wait no more, so just start the show. First, a quick word from our sponsor. Luis, we are so excited to have you on today's show because we haven't just been researching

yet for weeks. We've been covering your company for years. And now we get to actually speak with someone who is the brains behind it. So you founded Duolingo and I p.oed the company in 2021. 40 million humans use it every day. 10 million of them are paid subscribers. You're growing 40% a year right now. The stock is more than doubled in the past two. years. The company is worth more than $13 billion. That's like two lifts. Duolingo is the most exciting education brand on the planet. Now, Duolingo at its heart, it begins and still is a tech company. And you've said the majority of your employees and your engineers and your designers, they are tech people, same as every other tech company. But unlike every other tech company, you've chose

to launch, grow, and scale yours in Pittsburgh, Pennsylvania. In fact, you've said you would never opened an office in Silicon Valley, the home of tech. Why is Pittsburgh your headquarters? Well, we were here and we wanted to stay here. I mean, I was here because I was a professor Carnegie Mellon. Severn was also here because he was a PhD student here. And so we were here and we decided to, we saw no reason to move at the time. And then we just started growing and growing. And at some point, we were kind of too big to move. In retrospect, you know, a lot of people have asked me what would have happened had I started Duolingo in Silicon Valley. And of course, it's impossible to tell what really could have happened. Like, I don't know. There's pros and cons of having

created it here. Some of the good things are, one is the people that you hire here stick around for a very long time. In fact, if you look at right now our employees, I mean, the average tenure is pretty high. I mean, a lot of people have been around for a long time. That's a good thing because you get these people that get really good at what they do because they've been doing it for eight years or however long. So that's a good thing. The other thing that I think was good was in terms of hiring executives, it allowed us to hire we punched above our weight because we ended up hiring an executive team

many of whom for one reason or another had had long careers for example in Silicon Valley but had to be here in Pittsburgh either because they had aging parents here in Pittsburgh or something and they had to be here for one reason or another and we were the best game in town

and so had we been in Silicon Valley they wouldn't have worked for us they were kind of outside of our league right so that this allowed us to hire really good executives That's another thing. Bigger fish in a smaller pond. Yes, we were the biggest fish in this pond.

Yeah. That was good. That helped us. And then the last thing that helped us, I think, is, you know, we were, when we started in 2011, education has never been hot in terms of, you know, funding or whatever. It's just what it allowed us to do here in Pittsburgh is it allowed us to not get distracted by the latest fads. I'm pretty sure that had we been in Silicon Valley, we would have quickly, at the time, it was like hyper local apps

and then it was like social apps and then it was like crypto and then it was like I'm pretty sure that we would have gone through all of this and at this point Duolingo would be like a I got an AI SaaS platform we're side hustling outside this podcast

right now Louise And you know it's not that it's that obviously Silicon Valley's amazing there's all kinds of good things about it but for something like education I think I think you want to stay for a long time just concentrated on what you're doing

and growing it and growing and growing it And I think being outside of Silicon Valley helped us for that. Of course, there were many downsides. I mean, sometimes hiring here was slow. I mean, we couldn't hire. We couldn't find the right person. Most of the people that we've hired to work here in our Pittsburgh office, we've moved from other cities to work here,

which is why we've opened an office in New York that is, I mean, we have a few offices, but we have a relatively large one in New York, and that actually helped us hire quite a bit. We're pretty happy with our location strategy at the moment, which, by the way, our location strategy at the moment seems to be to have offices in grave, places. Not great, great.

Like Seattle, we have an office in Seattle, we have an office in Berlin, we have an office in Detroit, we have an office in Pittsburgh, we're masters of great places and that was not on purpose, that's just what ended up happening. You guys look at like the percentage of precipitation every year. We don't, but we may as well, we may as well, it really is, and we also have an office in Beijing, which is great for different reasons because of pollution. But basically, we have an office in great places. So very quickly we became the most downloaded app in education. This is even before we became iPhone app of the year. We were the most downloaded app in the education category. I just didn't know we were here to stay, but we became the most

downloaded app in the education category. And it was because it was free and it was fun, and you could actually learn some languages with it. And so that made us kind of stand out, mainly that it was free and fun. And so we were the most downloaded up in the education category. But the next big thing that happened was around 2016, 2017, we started monetizing. Before that, before 2016 and 2017, we were not making any money. Duolingo was just free. There were no ads. It was just free.

There was no way for us to make money because we didn't have ads in the platform. We didn't have subscription, nothing. We started to make money in 2017. It was actually hard to make money. Not because nobody wanted to pay. It was internally our employees didn't want to make money. Like mission-oriented? Yeah, no.

We had hired. Everybody that we had hired was very mission-oriented. And then when I told the company, it went to figure out a way to make money. money. Everybody was like, why? And I'm like, well, let me tell you, we need to pay your salaries, for example. And they would say, but you've been paying our salaries this whole time. What's the problem? And, you know, of course, what we had done is we had raised venture capital more and more venture capital. You can't raise venture capital forever. And so we started trying to figure out how to make money, but it took about six months to convince the employees that it was okay to work on this. In fact,

the people that we had chosen to work on this were kind of boycunning the work. They didn't want to work on it. They were like, I don't really want to work on that. Eventually, what worked is that we were able to convince everybody, and I do believe this to be true, we were able to convince everybody that, A, making money is not evil. And B, that if you make money, you can use that to reinvest in your mission and actually grow a lot more and really fulfill your mission a lot faster. Scale your impact. That is true. And in fact, that's exactly what happened. As soon as we started making money, we could scale our impact.

But I think that the big part of this is when I started seeing that it ended up being relatively easy to monetize Duolingo. I mean, it was hard for the employees. But once we decided to do this, just every year we started making more and more and more and more and more money. And that was a big boost for us. Well, Luis, Jack and I have been keeping a list on the side here because what also catapulted your guy's growth was that you're a psychology company. The way we see Deolingo is psychology. You're a master of mind gamification. I mean, you must have a bunch of psychologists on staff

because the push notifications, the streaks, even the mascot, they all tug at just the right psychological courts to get people to open the app. Let's talk about the psychological growth hacks that got duolingo to 40 million people opening it every day. What was the first psychological engagement hack that you employed? The main thing that has set us apart, and especially as an education company,

is that our product is engaging. And that is the most important thing. It is also, I believe, the most important thing, if you're trying to learn something by yourself, the hardest thing is staying motivated. So what we do is teach you. And sometimes people get confused about this. People are sometimes like, well, did you know you can learn whatever it is,

not just languages. You can learn whatever on your phone. And I'm like, that's not impressive. You can learn anything on books. Like the reality is you can learn things. There's been hundreds of years we can learn things by yourself. The hard thing is keeping people motivated. That's the hard part.

And the first thing that happened was when we were before we launched. Louise, can I ask a quick question? When you say it's engaging, how is your app, can you just like break down that word for a second? How is your app more engaging than a book? Well, people use it every day. If you just look at retention, as in the number of people that start and come back the next day, is much higher than a book, I would say. And it's also the word of mouth that we have.

Retention gives you scale. And we also look at how much time is spent by people doing it. Essentially, people get quite engaged. We don't like to use the word addicted, but it's kind of like, you know, many people you could say that they're addicted to it. So is your app addictive? In certain ways it is. I mean, it's extremely engaging.

It really gets you to continue doing it and use it every day. How did you get your app to be so addictive? Well, the first thing we did was to turn it as much into a game as possible. Okay. And this is something that came up even before we launched, and it was because we couldn't get ourselves to use our own app, Before it was a game. I was trying to learn German because Severin's native language is German, and he was trying to learn Spanish because my native language is Spanish.

I made the Spanish course. He made the German course. We were learning each other's language. We couldn't get ourselves to use it. Never go when you don't like the taste of your own cooking. Yeah, this was not good. So then we decided, okay, well, we need to do something about that. And this is when we decided to make it as much into a game as possible. So the first thing we did is, you know, lessons used to be 30 minutes long. We turned them into like three minutes long.

Wow. We had a progress bars everywhere. We started giving points. So all of that, it just kind of turned a little bit into a game. That is kind of step one. The next thing we did related to that, this is after we launched, but we added this notion of a streak, which was really powerful. A streak just counts the number of days that you've used the app consecutively,

and if you stop on a given day, it goes to zero. That has really been transformative for us, the streak. Now, with the streak, you would think that it's just a number that you put in there, and it is, it's just a counter, but it's actually extremely sophisticated. You would not imagine how sophisticated things are. For example, it's not quite true that you can never miss a day. We give you these things that are called streak freezes in case you miss a day. I've experienced, full disclosure, I've experienced this.

I didn't want to admit it publicly, but I'm willing to do so in front of you, Louise. I've told Jack before in private. We do that, but there's a lot of science behind when to give you a streak freeze. So as not to cheapen the streak, but actually get you to keep going. So we give you streak freezes. The other thing about the streak that is really important is the notifications. We cannot let you forget to do your Duolingo. And so we have notifications that tell you,

hey, you're going to lose your streak. And there's a lot that goes into what we should say and when we should say it to you so that you don't lose your streak. And at this point, we have a pretty sophisticated, it's basically mostly chosen by AI, you know, when to send the messages. But we also have a very quirky voice. The messages that you get from Duolingo are pretty quirky. and we've A, B, tested them like crazy

and they're personalized to you in terms of what works to bring you back. So for some people, we use guilt. For some people, we use different things, and then it really works at bringing you back. So the combination of these two. And what does that look like? What works for a notification?

What does a guilt notification look like? What is the best timing look like? What's optimal here? How are you guys doing it differently? It ends up being actually quite complicated. But I can tell you some things that we've learned generally. Generally, the time to send you the notification is about 23 and a half hours after you last used to a lingo. Okay, why's that?

The way to understand that is if you use dualingo today at 3 p.m., it probably means you're kind of free at 3 p.m. And likely, you're kind of free tomorrow at 3 p.m. too, unless it's a weekend. Weekends are a little different. But weekdays... No, you're right. In terms of what to say, I mean, we have found all kinds of things, but, you know, one of the most powerful things. We found this many years ago. So we send you a notification, and then if you don't use duolingo, we send it to you the next day and the next day.

after five days of inactivity, we stop sending you notifications because we think we're not going to spam you forever. One of the things that we found is, you know, we thought, well, if we're going to stop sending you notifications after five days, we may as well tell you that we're stopping. You know, we're being nice to you. We're not going to spam you forever. We may as well tell you.

We didn't think this was going to bring people back. But what we did is we started sending the last notification that we sent says something like these reminders don't seem to be working. We're going to stop sending them for now. We thought we're just being nice. I didn't realize... Yeah, you're being honest. Yeah, but I didn't realize

this was an extremely passive, aggressive way to get you to come back. Yeah, a little side-eye. I'm picturing the mascot saying this. It works really well. It works really well. And that notification,

not only that notification worked really well, it also kind of became the basis of our voice. Like the brand? We started really realizing, okay, we can be a little more savage.

I mean, this is what, before that, it was all very technical. Like, no, this is when We started, and also that notification started giving rise to all these memes about how the owl, you know, really wants you back. And that's all the owl wants for you to do the language. And he's willing to go through great lengths, including kidnapping your family or whatever it is for you to do your lesson. But it all came from that notification. And then there's all these memes that came about.

And our voice really was found. And at this point, we have a pretty quirky voice that is, yeah, that is a little savage. It makes fun of you sometimes. But it generally works to get you to come back. We got to ask you about The Owl because we have a long list of questions about The Owl. But Jack uses Duolingo to keep up with his German, which he studied in college. I used duolingo, particularly like before a trip to Italy to like catch up on my Italian in the month leading up to it. So, you know, like I'll be walking down in the street and someone will hear me say, I am a potato.

And like, I'm actually using duolingo, but I'm saying in Italian, we were curious, you know, in terms of the different languages people are learning and the cultures where they are learning these, are there differences in terms of what gamification tactics work for people in different countries or learning different languages? No. It's universal. We have heard so much from people who tell us that there are differences. And we also employ country marketing managers in every country, not every country, in every large country. And they tell us all kinds of things like, oh, you know, Brazilians, they're different for the following. And they tell us all this stuff. But people in China, they are different for this reason.

Nope. It turns out pretty similar. The things that work, work pretty much everywhere. Generally, and we just have not found things that are very different everywhere. There's a little bit of, at least not in the app. The app is the same everywhere. Our social media, you have to be a little different in Japan. Japan's a little different in our social media. But every other country is pretty much the same. So what are the words you use to describe the motivations that you try to get your users to have. Like, guilt is clearly one of them. You'll try to make your user feel guilty that they haven't opened the app in the last seven days.

Streaks about to end, Jack. Is there also an inspirational version of that push notification? Yeah, there's inspiration. There's what people want to accomplish. I mean, we have found different parts of the life cycle. Usually the first, when people just sign up to the app, they're very, you should be very inspirational. And they're usually very much, like, they have dreams about what's going to happen. Then reality hits after a few days when you realize, oh, this is going to, it's going to take me years to learn a language, because the reality is it takes years to learn a language. And this is when you start having to apply other things. But yes, you know, motivating messages helps.

Friends, you know, of the last few years, dual language has become a lot more social. Social connections and social pressure helps quite a bit. And also completionistic stuff, like you've done the last five days, are you going to really miss the sixth one? Like, kind of completionistic really, really helps. Do you have a network effect or social proof? Yeah, we do. I mean, by now, Duolingo, about a third of our users use our social features quite a bit. Those are people that have at least three mutual follows, so three people that they follow and follow them back. And then, you know, on Duolingo, you can send congratulations. You can have friend streaks, so free a streak together with your friend, et cetera.

But a third of our users use this pretty heavily, and that has network effect. Before we jump to the mascot, Nick, I just want to point out, I found it wild that Louise volunteered the word addicting. Yeah, I know. I know. I was thinking the same thing. Because like 39 states attorneys general are going after Instagram for employing addictive features to their app.

I think the TikTok company does not allow the word addictive to be used internally or you're out of there. It's Voldemort over at TikTok. However, you're broadcasting it to the world that you're using addictive gamification strategies. It's hard to say that these are not. However, what I will tell you is that we really are trying to get people. I mean, our ultimate goal is to teach people something.

And if you think about it, what we're doing is we're using the same techniques that are used by social media apps or mobile games, et cetera, because you have to use them. Otherwise, you've lost the competition. I mean, there is a major competition for attention. We ourselves, you know, what I tell people is our competitors are Instagram and TikTok and mobile games. That is who we compete with because we compete for time. You're saying it's not other language learning apps. We notice you're 12 times bigger than your next closest competitor. They're way behind you.

If you look at active users of language learning apps, we have somewhere between 90 and 95% of them of all active users of language learning apps. So when we see that somebody leaves Duolingo, they don't go to another language learning app. They just spend more time on Instagram. That's the reality. And so our competitors are TikTok, Instagram, Snapchat, all the places where people are spending time. And they are heavily using all of these techniques. If you don't use them, you're fighting fire with a squirt gun. You're just not a successful app these days.

You're just not. I don't want to go full Game of Thrones here, but it feels like you guys are saying you're using the same weapons as the enemy, essentially. You're using dragons as well. You're just using them as a force of good, not a force of evil. I mean, if I could make broccoli addicting,

I would find a way to do that for my children. If my teachers made reading books addictive, I would love that. I'm on board. You're saying addiction doesn't have to be a bad word, Jack. I mean, that's how we feel. And, you know, again, obviously there's all kinds of things. But, you know, dual-lingual, we're far from what would be like a TikTok or whatever.

We're, you know, not only are people learning languages, the realities people spend a lot less time there, etc. So, Luis, in a world now where you see people saying less screen time, no screens, anti-screens for kids, do you feel like that's a threat or do you feel like there's an exception here that should be carved out for something like an education app? I don't think this is a legitimate threat. Fair. People say that. that's not what the metric say. Could you share more?

I mean, people may say that, but the reality is that this is kind of unstoppable. It's the same as people saying kids shouldn't watch TV. I mean, like many years ago, I don't know. When I was growing up, TV was rotting your brain. And you shouldn't watch TV. I don't think that worked. I think people watch TV. In a similar way, I mean, I think this is just what the future is.

This is the present even. The genie's out of the bottle. Yeah. So it's an operating principle at Duolingo that people will be on their screens all day. I don't know if it's a... We have these other operating principles that are actually operating principles, but it is a belief that people will be on their screen. That's just a belief.

Let's talk about another product you produce that transcends global boundaries, borders, barriers, and culture. Which is your mascot? Let's talk mascot strategy here. It's the least techy thing about Duolingo, the mascot. But it might be the most culturally relevant thing that you've accomplished. the unhinged duolingo owl duo. How did you choose an owl?

We chose an owl because it is, at least in Western cultures, a symbol of knowledge. Turns out actual owls, probably not all that smart and they're vicious. But that's a symbol of knowledge. But what made you think as a tech company, we can't think of any tech companies that have a mascot like a sports team. But that's essentially what it is. This is a costume people put on in the office. What made you think you needed a mascot in the first place? Very early, before we launched, we thought that having a mascot would make us friendly as an education company.

And so we decided to have a mascot. Then we decided it was going to be an owl. Then we decided it was going to be green because Severin, my co-founder, hates the color green, and we thought it would be funny if it was green. So we've had it all along, and a lot of it, you know, it became so popular, a lot of luck. I mean, we did some stuff. We kind of found our voice that it was a little unhinged. it was a little gil-trippy, et cetera.

But it was a little luck that all these things happened. For example, the fact that the memes started with the owl, we didn't start the memes. What we started was that notification that said, it seems like these reminders are not working. We're going to stop sending them. That got the memes to go. And that was not us. Then we ourselves leaned into the memes. We saw that that was happening.

So we started making fun of ourselves. Embrace the meme. We did. And so we produced, for example, we produced an April Fool's video. a lawyer's commercial. You know those really shitty lawyer commercials that you see on TV? Oh, 1-800. Get law now.

Yeah, so we made one of those, but it was a lawyer saying, if your family has been hurt or kidnapped by an owl, I would get you money. What I love about this is that it has nothing to do with learning a language. This seems completely unrelated. Not entirely.

It's not necessarily learning a language, but it's learning. Look, the best characters out there, They all have something in common. And it is that they want something and they have an obstacle to get it. Duo, the owl, the only thing he wants in life is for you to do your lesson. Wow. And the obstacle is that you're lazy.

So it is related to learning. I mean, we really, that guy just wants you to do your lesson. And he will go through great lengths to do that. That's what makes them a good character. But a lot of this, again, a lot of this just kind of evolve through time. It's not like when we started the company. We're like, and we're going to have this owl that's going to, do crazy dances on TikTok or whatever.

So there was that. The other thing that happened is a few years ago, already the owl was pretty popular, but a few years ago, we started a social media strategy that honestly started despite me. I didn't want to do that, which was doing these videos with the owl suit.

I thought, the first one I saw, I thought, this is pretty low production. We spent so much effort on an app being very high production. Like millions of dollars on that being high production. And then we saw these videos that were like super low, low fidelity and low production. I'm like, what is this? But it turned out people really, really liked this.

And that has made it super popular on TikTok. At this point, our owl is so popular that it internally, it has an agent. The mascot has a talent agent. Yes, internally. Like, we have a person in our marketing team. There is an agent. But what's amazing about it is it's so popular now that famous people want to interact. with the owl without us paying. So we just get all these famous people wanting to like do

co-labs with the app, with the owl. Because sometimes the owl has more followers than they do or their videos become, our videos become more popular, et cetera. So it's been a pretty big deal for us. Yeah, probably was the reverse of how it started out. You would have paid to be, you know, have Duo the Owl, hanging out with Duolip of the singer. And now it's vice versa. She wants to date duo the owl. The old switcher. It really is this thing. I mean, before people were coming to us to try to pay them to hang out with us. That is no longer the case. Luis, you said that your employees early on

were very mission-oriented. I'm sure they still are today. They still are. You did that TED Talk, and you told us that wealthy people pay for the Duolingo app so that poor people can use it for free. Why is Duolingo the rare app

used by both migrants and MBAs? Yeah, I mean, the first thing is that, you know, it's free to use Duolingo. You can use it, and you really don't have to pay us. We don't really block you from anything. or anything, you just use it. And that has made it so that it's really the whole, the whole socioeconomic spectrum uses Duolingo. Like Syrian refugees in Europe are using Duolingo to learn the language of the country that they relocated to. At the same time, on the other side, we have a bunch of famous

people or billionaires that use Duolingo. And it really is because it's accessible to everybody. And this really is what we wanted to achieve. We wanted it to be the case that if you had more money, you couldn't get a better system. It's the same system as being used by people who have, you know, no money versus billionaires. So that's usage. Now, we have to make money somehow. And it turns out the way we make money is quite compatible with this, because the way we make money is we have this freemium model,

which is if you don't pay us, you can use tooling as much as you want, but you have to see an ad at the end of a lesson. We make some money from the ads, but if you don't like ads, you can turn off the ads by paying us to subscribe. Turns out a small fraction of our users pay us to subscribe. So if you look at our monthly active users,

only about 10% of them pay us to subscribe. But this 10% gives us 80 to 90% of our revenue comes from this 10%. So we don't make that much money from ads. We make most of our money from this description. As contacts with like Netflix or other subscription services, it's a much higher percentage that are paying. Much higher. You just, yes, that's where the money comes from.

Now, what is interesting for us is because the app is freemium, what ends up happening is that people who don't have a lot of money just don't pay. The people who pay are usually relatively wealthy. And I'm not talking, they may not be millionaires. But, you know, people with stable jobs, in well-to-do countries. That's who pays us. And the people who don't pay us

are usually people who don't, may not have a job or who live in a poor country, like, you know, in Brazil or Vietnam or something. So what we like about this is that it really is the case that the wealthy people are paying for the education of everybody. And they probably don't even realize it. Probably not. But it is something that we like a lot.

Have you ever thought about making these wealthy users realize it? Yeah. I mean, we tell them, actually. Sometimes we say by, you know, subscribing you're helping support free education for everyone. So we say that and that does get people to buy, but the main thing that gets people to buy is turning off the ads. Let's be honest. Have you noticed different characteristics between those using your app to learn English and those learning your app for other languages?

Yes. English is a unique language. It really is very different from all the other languages in terms of the desire to learn it. Some people sometimes say that they worry that Chinese is going to be like that. I can tell you we have the numbers now. English is a unique language that everybody wants to learn, not very many people want to learn Chinese. The reason is unique is because it is the one language we're really knowing it makes you more money. Most other languages, knowing them, unless you want to move to that country, but generally, it's usually a hobby. So generally, when we talk to people that used to a lingual that are learning any language that is not English, you ask them why are that you're learning the language?

The most common answers are like, oh, I used to play a lot of candy crush. and now, you know, at least I'm learning a little bit of Spanish. Like, that's that. But for English is different. When you ask them why are using the app, they're like, no, I need to learn English. Wow. And I need to learn English because I need to make more money. So English learners, people who are learning English, are just dramatically different

from people who are learning other languages in every aspect. Is English the top language in every country? No, not in the U.S., for example, it's Spanish, because most people already know English here. but in non-English-speaking countries, English is the top language to learn. Yetis, we're about to take a short break, and then Luis will announce Duolingo's next big product. So, Luis, we know your ambitions are so much more than language learning.

2022, you launched math learning. 2023, you launched music learning. We're excited for you to announce your next big product right now on the best one yet. Yes, I'm very happy to announce this. Like you said, we don't doolingo. you know, you can learn languages and math and music. We're adding another subject, and it is not one that many people have expected,

and it is that you will be able to learn chess on Duolingo. Chess? You know, it's a game, but it is, you know, it's an after-school program, and it is a game that really can make you smarter. We know there's just a lot of people who will tell you very similar as with Spanish, where it's just like, I've always wanted to learn chess,

but I kind of never really had the time or seems intimidating, et cetera. So our goal is to get a lot of people into chess. and the idea is to let them learn how to think logically. One of the reasons why we started, there's a number of reasons why we started, but one of the reasons I greenlit the project, by the way, was not my idea to teach chess.

But I green lit it. And one of the reasons was because in my country in Guatemala, our education system in Guatemala is just not working. One of the things that the Minister of Education told me was, look, we're failing so much that at this point, I just want them to learn how to think somewhat. So you know what we're going to do? We're going to start teaching them chess.

Because it's at the very least they learn how to think logically. Learn how to think. And that's the thing about chess. It lets you learn this kind of if-then-else statements. If I move here, you will do that, and then I'll do that. And so it just really gets you to think logically. And so we're pretty excited about it. We also think there's a large audience for it.

I don't have a very precise date for it, but it's coming pretty soon. Yeah, this is an amazing project. It's all kinds of good things that have happened since we started working on it. Huge congratulations, Luis. Jack and I know how much goes into a product launch. We always say ABLE, always be launching everything. Can you tell us how this began beyond that insight from Guatemala? Like internally, as a team, how did you guys build this?

How did you choose chess? There's a number of things that happen. But in terms of how we chose chess, we're looking for subjects to teach that have very large audiences. I mean, hundreds of millions of people want to learn them. Chess, I believe, is one of those things. We also are looking for things that will somehow make you smarter, that we actually really want you to improve yourself. And we're also looking for things that take years to learn.

We're not so interested in things that you can learn in like a month. Then we'll lose you as a user. So we want to do years to learn. And chess is something that you can really get better at for years. So you're looking at for something where there are a lot of potential players or a lot of potential users where there's an opportunity to learn and it's going to take a long time to learn. This is the things that we're looking at.

And one other thing that we can do a good job with a mobile product. Okay. Yeah, so swimming wouldn't work. A lot of people want to swim. And chess is perfect for a mobile. So all of those things were true. Then what happened was we have a person internally who we love. His name is Tyler, Tyler Murphy.

He's been here forever. And he's started a lot of really our major initiatives. Tyler came to me and he said, I want to do chess. And, you know, I went through the list and I'm like, it satisfies everything that we want. Checks every box. Let's do this. And so Tyler started working on chess with this other guy, Edwin. two very long-term employees, long-time employees of dual-lingo.

And neither of them is an engineer. Tyler's a designer and Edwardine's a product manager. And what was amazing about chess is all other projects that we started in the past, we needed to put like a team on them with engineers and everything. In this case, we did not. We just let them, we're like, okay, you know what? At the moment, I have no engineers to give you. You figure it out.

Side hustle. And what they did is they basically started messing with large language models. and for the first several months, they made an entire chess course through large language models, no engineering required. And they came to me with prototypes that they had made,

even though these people were not engineers, of how to learn chess on Duolingo. And it was just the two of them, and it took them a few months now. Of course, by now we're about to launch. We now have a team that has actual engineers to put the stuff in the app, et cetera. But for the first several months,

normally we would have put a much larger team. It would have taken us much, much longer, but this really is a case that with large language models, we're able to do this really fast, actually. So two non-engineers plus AI have led to the next big duolingo launch, chess, that you're now all in on. 100%. That is what happened. Again, I do want to emphasize, we now have engineers and a team and everything, but this is what happened for the first, call it, six months. How big a bonus are you going to give Tyler an edit?

Why'd you ask that? Now you put it because I know I know these two dudes are going to be listening to this podcast and now they're going to be like Yo, where's my bonus? Why'd you do that? Jack, I think I see the owl

showing up behind Luis with a checkbook. What is this? Don't do that. Don't do that. Anyways, they deserve huge bonuses. They're amazing and what they did was amazing. Very excited for this too

because Jack and I were college roommates and I believe we started a game of chess once Jack where like you moved a rook and then I asked you what rook meant and then I moved to bishop and you said, you can't move the bishop like that. I think we never finished that game, so I think we're overdue for this thing.

It's pretty clear you did not know how to play. But now you can learn on duolingo. I need to get a streak, Ellen, Louis. Called the castle. I'm telling you, what's amazing about, you know, I've been using it, of course. It really gets you from zero. Not only do you learn how to move the pieces,

which is important, but also you learn strategy. And you get a pretty good spot. Also, a great time to launch a chess product. I feel like the chess world is getting disrupted right now by technology. There's that new, like, alternative. to the World Chess Championship that does random chess,

where the board starts just randomly, not the way it's supposed to? And Jack, we've covered other chess companies that have gone viral since the pandemic, which are a little bit more advanced, I think. I think those are focused on, like, your advanced players who want to play more complicated games. So this is kind of like,

it looks like you're targeting like an entry point, Louise. This is it. I mean, eventually we'll probably do higher levels, but at the moment, we're targeting entry point. I mean, if you either don't know anything about chess or roughly know how to move the pieces, but don't feel very comfortable playing chess with anybody because you think you're going to get beaten.

This is a great product for you to use. And that's a lot of people. It should say one other thing, which is really interesting, and it's a big belief we have a Duolingo. You know, the first 20 employees of Duolingo, we're not very good at learning languages. We were not language nerds. We were, none of that.

I'm not very good at learning a language, and I'm not a language nerd. I'm not particularly interested in learning languages. And what that did is it made us create a product for the average person. Had we been language nerds, we would have made a product for language nerds. Interesting.

And I think this is true, and by the way, Tyler and Edwin, they are not excellent chess players. I am not an excellent chess player, but I'm pretty sure I'm significantly better than they are. These people, I think a few months ago,

we're still kind of figuring out how the night moves. But it's amazing because I really do think they are making a product for the actual beginner. You see Luis knocking Tyler and Edwin down? a peg or two for that. No, no, no, no.

Love these guys. Love these guys. I'm not saying anything they don't know, okay? It's channeling owl. He's channeling the owl. You know, Luis, we were lucky to have a pre-interview chat with you. And yet he's just, you know, when Jack and I do these interviews, we like to speak with our guest a week before, so we get to know them and really kind of get a sense of

everything they want to talk about, too, and what we want to talk about. And in our pre-interview chat last week, you told us that artificial intelligence was a huge area of focus. And you also just revealed there was basically the third. third player launching your latest product. Now, honestly, the way Jack and I see artificial intelligence being in the media business is it can be both a threat and an opportunity at the same time. First, the threat, Bloomberg reported in March that Apple is working on real-time language translation built into AirPods powered by AI. So that would look like a threat to a duolingo.

At the same time, AI allowed you to launch AI tutors, which is your new premium product. You can practice your German with an AI. What's her name, Lily? With Lily, yeah. That sounds really cool because Lily's not going to judge me if I misconjugate my German verbs, which you wouldn't, Jaggy, you're doing fantastic. You got a 12-day streak going.

So we've got to ask, is AI a threat or is AI an asset? Oh, for us, it's a major positive. Okay. From the beginning. And this is, by the way, it's the last couple of years that AI, it's like AI is the next crypto. It's like, it's a big thing. But, you know, when we started duolingo a million years ago,

AI already existed. It's not that the large language models, but we were big believers that it will be a computer that teaches you. And I think in the future, more and more this will happen, that the primary way of teaching is going to, or learning, is going to be learning through a computer. To your credit, Luis, we looked at your IPO paperwork before this interview,

and this was before ChatGBT. Yeah, 100%. Well, first thing, first thing, we did notice a typo in your IPO paperwork. No. No, I'm just kidding. I'm just kidding. No, no, no, no. You don't understand how many times that was read. You don't understand how many times that was red.

If there was a typo, shit. No, you don't have to fire Morgan Stanley. No typos, we promised. But we did go through your IPO paperwork. What did we notice, Jack? We control after it, and we found artificial intelligence mentioned eight times. So, yeah, you were thinking about it before this virality. We were.

And we've been for a long time. I mean, the reality is we really have thought that a computer can teach you. We still believe that. It is the, you know, a computer head. There's so many good things about a computer teaching you. It's infinitely patient. It's way cheaper to teach you to a computer. computer, so kind of things, especially, and if you're practicing conversation, the beautiful thing is

you don't feel bad practicing conversation with a computer in a language that you're not very comfortable with. But if you're practicing with another human, a lot of people are shy about that. And so this is, this is just, it's just better. Nick, I was a German major. And then I worked at a German bank after college. Yeah. And sometimes I had to speak to German clients on the phone. But I had a cubicle surrounded by my peers. And every time I had to speak German, I would book a conference room to make sure I was in a private place because it's embarrassing. It is just this. I speak Italian sometimes.

I'm not speaking to this volume when I'm speaking Italian. Right. The same thing happens to me. And when I speak a language I'm not very comfortable, I mumble it. Because I'm like, me. Yeah, yeah, yeah. Because I'm shy. I'm shy about it.

It makes you feel foolish. Yes. But when you're talking to a computer, you don't feel foolish. So I really do think it can be a positive. You know, in terms of the threats, I, you know, people say all kinds of stuff. Like, for example, you know, like you said, there's going to be real-time translation. Why do you need to learn a lot of it? language. The reality is that real-time translation has been around, you know, Google Translate has

worked really well, and it has been very fast, and it's been around for a long time. It's actually quite useful for things like traveling. If you don't really want to learn a language at all, and you're traveling to a country that for some random reason, nobody speaks English, which, by the way, these days, is very rare. You can use this to kind of get around real-time translation, and it's great. But if you look at our users, the two buckets of users, one is people who are learning as a hobby. I don't think they're going to be affected by real time translation. It's a hobby. It's the same reason why people learn chess, for example. You know, computers have been really good at chess for at least 20 years. People still learn chess.

So it's this hobby. And then the other big group of people that are learning a language here are learning English and they actually want to learn. It's not going to be good enough for them to be walking around with airports all the time. They actually want to learn English. Luis, there was one word you've mentioned before that Jack and I were fascinated by. We were shocked you used because we haven't heard any other techies use it, maybe because we're with so many in the Bay Area. But you said the word artistic. You employ literally hundreds of artists at Duolingo. They craft the branding, the imagery, the clever notifications that you mentioned before. We were curious about the artists. I really believe that this is something that has differentiated

Duolingo. We call it the design department. It's mostly art. There are a lot of artists in there. About 150 people in there. Wow. By the way, bear in mind, this is a company, the entire company has 900 people. Wow. So about 150 people are, you know, artists. Art majors. And they really, it's amazing because they come up with all the unexpected stuff. They're the ones who come up with, for example, we did a Super Bowl commercial last time, last Super Bowl, not this one. What it was is it was, it was, it was, he was sitting backwards. You could just see his back, but you could see his butt crack. And then it started growing, like his butt started growing. And then suddenly what it was, it was a Another owl grew out of the owl's butt.

I can't come up with that. And then it just said, do your duolingo. You need an art degree to come up with that concept. Who comes up with that? This was, you know, internally, internally our artists just came up with that. And it made no sense and what the hell. But it was, you know, it was one of the most talked about, it may have been the most talked about, or one of the most talked about Super Bowl commercials.

It was only five seconds. And it was just the owl, the old butt growing into another owl. But this is an example of the stuff. they come up with. They come up with really weird stuff. And I actually believe that as AI matures, what's going to be most valuable is taste, you know, because you could just do all kinds of things. For example, coding is getting easier and easier. And so probably a lot of the engineers are going to become kind of conductors of coding. And I think just having all these artists here is very valuable because they have a lot of taste. I mean, they have, they really kind of have

an idea of what's weird and what's quirky and what's unhinged, that lets us go move the brand forward. If the AI revolution makes us all more artistic, that would be fantastic. Well, I don't know if it'll make us more artistic. I think the artists are going to do better. And by the way, there's something that's something that one of our engineers, I was telling one of our engineers who pointed this out. Before like 50 years ago, math nerds were not in a good spot. They were just, they kind of were ostracized. At some point, because of technology, math nerds, like the richest people in the world are math nerds. And like, they got a boon.

It's likely that because of AI, math nerds are not going to be all that valuable anymore. It's cyclical. So basically, we had a good 50-year run. Like us math nerds, I am a math nerd, we had a good 50-year run. It's probably over. And it's going to be back to the artists. This leads to a big question that the Yetis and the besties,

our listeners, ask Louise, which is, we want to be the best at what we do. How do we focus our careers? You know, a lot of them are business school students. or they're working in marketing, or they work at a bank, or they have a job where they're curious about the next five years,

and they're maybe worried about the threat of AI. So what advice would you give people in the age of AI? Look, I don't know, but I'll tell you one thing. It is our belief of dual-lingual, you know, I've told this to people. Because sometimes people here ask me, you know, our computer is going to take my job. What I tell everyone is,

a computer's not going to take your job. A person that can use AI will take your job if you don't use AI. So the one thing that I'll tell you is, no matter what, do not stick your hand in the sand. and not you say, get as good as possible with these tools as you can, to do everything you can with these tools. And that's what we're telling people internally.

That, I think, is a no-brainer. Like, you should just do that. Now, in terms of what to concentrate on, that's very hard to say. You know, it's almost like, we asked you earlier, is AI an asset or a threat? And you, without hesitation, said asset. Yeah. I feel like your advice is almost,

people need to get to a place where they have that same response. Yes. They need to tell themselves it is an asset. That's step one, accepting that it's the asset. But it's inevitable. AI will come and you cannot stop it. And it is coming and who knows at what speed and who knows, but it is inevitable. So best thing you can do is jump on that. Make AI an asset so that you're better at your job. I thought you were going to say make AI great again. Another way to put it, if you have the opportunity to go to MIT or RISD, Duolingo thinks you should go to RISD. That's your words, not mine.

One thing Jack and I noticed in our research, thinking of you guys as a psychology company, is that you guys as a psychology company, is that you know, intrinsic versus extrinsic motivation. How do you guys think about that and how do you employ it? And what is that exactly? Intrinsic motivation is your motivation to actually do the thing. And extrinsic is, you know, you're getting rewards for doing the thing, for example. Most people are very lazy, even the motivated ones. This is something we embrace a lot of dualingo.

For example, one of the things we do is in all our screens, there's very little text in them outside of what the actual exercises, because the mantra at dualingo is people. people don't read. It's because most people are very lazy. Yeah. And so our belief is that there's some intrinsic motivation that humans have for doing things, but it's not large enough to get them to do the things that would really benefit them. And this is why we spend a lot of effort trying to get them to motivate them to learn. The reality is most of us are motivated to do things that give us quick dopamine hits like TikTok.

If it's not fun, it won't work. Yeah. great quote of viewers that we noticed. Luis, like we mentioned, we dive in deep T-boy style, and we've pretty much seen and read every single interview you've ever done to prepare for today's show. But the one quote you've said that we love the most is that one. It's, if it's not fun, it won't work. We believe that actually with our show. There's an obvious value for news and staying informed, but people's intrinsic motivation only goes so far. So we try to make our show fun. But if learning has to be fun to survive and news has to be fun to survive,

Is there a place in this world for not fun things? You know what the problem is? We got scientific and we got really good at figuring out how to take your attention, not just, not duolingo, like humanity. Like, you know, a hundred years ago, it was a book. And then we're like, we discovered TV. TV is just way more engaging than books. Okay, good.

Then we discovered, like, social media is way more engaging than TV. The thing is, and we get more and more scientific and better at keeping your attention. I think, unfortunately, the boring stuff is gone because there's just so much amazing things that just keeps your attention. We just got scientific about it and we got good at it. And so at this point,

my sense is you just have to really fight for people's attention and you have to be using the techniques that are now best understood it. It's kind of a similar, you know, you could have said TV is really bad, but when I grew up, there was the Discovery Channel

before it turned into like reality TV. It was like science. But, you know, they were trying to use the same medium that all the other stuff was but for teaching you. It's kind of a similar. thing that we're trying to do here with Duolingo.

So any advice for our Yetis and besties in the audience who are building something to make something fun so that it'll work? Well, the first step is that you should like it and people around you should like it. It's hard to make fun things. It's hard to make fun things, first of all. The second piece of advice that I have is, and I like to tell this to a lot of people here at Duolingo, I have never seen a successful product that got really successful. in under, like, call it three to four years.

Wow. Of work. I mean, you may not have launched it, but everything takes, good products take, you know, three, four years. And this is a fast ones. You just need the time in the oven. And this is what I use this, I use this line here internally. You just need the time in the oven. It's like lasagna.

It just takes a while. It just takes a while. And no product has had overnight success that remains. As in, if it's long-term success, if you're talking about, you know, whatever products, Instagram, chat GPT, all of these things that have a lot of success, they were years in the making. I believe this is known as the Lazzania principle, Jack. Luis's Lazzania principle. It just takes a while.

It just takes a while. You got to leave in the oven a bit. Well, Luis, you got an inspirational story. Your company has done an amazing thing, which is make learning fun. We're addicted in a good way. Good, yes, in a good way. For something useful. We think it's fantastic that you're fighting the fire of TikTok with their fire tactics.

And it's just been amazing talking to you. Congratulations on all your success. And we're thrilled that you announced the product of chess on Duolingo on our show. Jack and I can finally finish that game, Knight to Queens Rook. You need to take back that illegal move, probably. I need a streak. I need a streak. Well, Louise, at the end of every story, we decide what the takeaway is. So we want to hear from you, Louise, what's the takeaway from you, the co-founding of Duolingo on Duolingo. Well, the biggest thing, I keep telling people, and nobody seems to listen to me, but I tell people. The hardest thing about learning something by yourself is staying motivated.

For some reason, there are a lot of education companies where, you know, at the moment, and of course, this can change the stock market and everything, but at the moment, we're, by a margin, the most valuable education company out there. And the main insight for us is not that we teach better, et cetera. I do believe we teach better, but the main insight is we keep people learning. And I think that's a big takeaway for Duolingo. Yiddis want more from Luis? Well, we've got an extended version up on YouTube.

We dive into his background, his childhood in Guatemala, the first company he gave to Yahoo, the second company he sold to Google, and more. So check out the best one yet on YouTube. Get the extended version of this entire interview. We got a link in the episode description. Jack and I, we'll be back with our regular daily show tomorrow. See you there.