上市了，就不得不向华尔街低头吗?

Hello. Hi. Good to see you.

Good to see you again.

It's funny because I just got performance review.

You got a performance review?

I get a performance review and I just got it. I can tell you what my

got it. I can tell you what my performance review says.

Foreign speech. Foreign speech.

speech. Foreign speech.

In most countries in the world, the brand of Duallingingo is really equated with language learning. So when somebody thinks about language learning, they just think of Dualingo.

Opening X because of AI. I really believe that the way people learn is going to fundamentally change over the next few years. Um, it just is the case that with

years. Um, it just is the case that with apps you can teach a lot better. So, I

made a decision a few months ago that we were going to prioritize teaching better and user growth over making money.

I made that decision.

That's big decision.

Yes, it is a big decision especially because we're a publicly traded company.

I knew that making this decision was not going to be popular in the stock market and it wasn't popular in the stock market for sure.

And so, our stock has gone down quite a bit. You know, this was on purpose.

bit. You know, this was on purpose.

because because we want to invest in the future rather than trying to make as much money as possible in the next 3 months. Is that like to gain more users

months. Is that like to gain more users to get more learning data so you can It's to get more learning data so that we can teach better. But it's also one of the things we've noticed is that whenever we get a user, we're very good

at keeping them there because the streak is there, all the gamification. So we

would just like to have a a much larger user base over time. We will make more revenue that way if we have a larger user base in the long term. Yeah, I see.

Generally, we've concentrated on teaching better, growing users, and revenue. Next year, we'll probably go

revenue. Next year, we'll probably go back to all three, but this year, we're mainly concentrating on teaching better, and growing users. It has been tough because the stock market has not been very friendly to us. But other than that, the company keeps growing very,

very fast.

You said like next year, you're going to do the three pillar thing to get the monetization back. What's the logic

monetization back. What's the logic behind? Is there something specific you

behind? Is there something specific you do this year or?

Um, well, first of all, one thing to clarify. It doesn't mean that next year

clarify. It doesn't mean that next year we're going to start charging. We're not

going to give features for this year and then start charging for them next year.

No, no. Whenever we give a feature for free, it'll stay for free forever. But

we we will concentrate a little more on monetization next year than this year.

As a company, we understand that it is important that we also make money. I

mean, we have to pay people salaries, you know, we have to make plushies. We

have we have to do things. Got. Yeah,

it's just like right now getting users are more prioritized. Yeah, I see.

You're going to show me something.

You ready?

Okay.

Nothing scary. Nothing scary. This is my streak.

891 days.

Actually, since we chatted last time, I started languages using Dualingual. I

think that's probably So, we talked last time 891 days ago probably. Yeah. Around that. Yeah.

probably. Yeah. Around that. Yeah.

Lily Dolly Eddie.

You're an education company, but you have a mascot, which is a green bird. Sometimes he goes crazy, right? This thing is like a

crazy, right? This thing is like a culture iconic right now for Dolingo. M

I'm just wondering at the beginning is it intentionally designed or is kind of grow organically?

It was both. We intentionally added a mascot because I always thought that a mascot would get people to come back. Oh

my god. See, he goes crazy. He He goes crazy.

Okay. Okay. Hi, duo.

He's going crazy. Um, we've always known that having a mascot is something that gets people to want to learn more. So we

put it in the first version of the app had the mascot and it was there. But

what happened more organically is his personality was in part invented by the internet.

It was by the internet. You didn't

design parts of it. We definitely designed that it was a green owl, but it didn't have much of a personality. The icon for the app is the

personality. The icon for the app is the face of the owl. So we decided that whenever the app sends you a notification, operating systems always put the icon of the app right next to it. So whenever you get a notification,

it. So whenever you get a notification, the icon is right next to it and it happened to be the face of the owl.

So people started thinking that the notifications were coming from the owl, not from the app. And so they started ascribing the personality of the notifications to the owl. At first the notifications were very simple, but we

started doing more interesting things with the notifications. For example, at some point we added a notification. If

you didn't come back for a few days, the notification said, "These reminders don't seem to be working. I'm going to stop sending them for now.

It's like have some personality, not just a notification.

And then people on the internet started saying, "Oh, the owl is passive aggressive.

The owl really will go through great lengths to do it." So then they started making memes about how the owl really will go through great lengths to get you to come back and do your lesson. Like

the owl would kidnap your family uh so that you do a lesson. It turns out that for characters to be good characters, what they need is they need to have something they really want. And for them to be a good character, the thing they

really want, there must be an obstacle for them to get it. In the case of the owl, all the owl wants is for you to do your lesson.

And the obstacle is that you're lazy.

Ah, so you figure people like this kind of personality and then you intentionally design all Yes. Whenever the memes on the internet

Yes. Whenever the memes on the internet would appear, we would look at it and we would say, "Oh, we should do more of that because people like that."

I see. So it's kind of like a iteration as well.

That's right.

The team sent me this this plushy. Mhm.

is quite big.

Oh, that one. Oh, that one. When that

was first suggested to me, I was a little No, I didn't say no. I was a little concerned.

Yeah, a little concerned. But my my chief of staff, it's a woman named Molly.

She wouldn't stop laughing for like a minute that we were going to have a boyfriend pillow. And so I thought this

boyfriend pillow. And so I thought this has got to be a good idea. Some of the marketing campaigns, I think, have pushed the boundary a little too much.

And we've sometimes said no.

It happened before, right? Did it ever happen? We've said no to a few things.

happen? We've said no to a few things.

I've said no to a few things. We never

actually did this. And I don't think they would have gone through with this one, but somebody suggested that the owl should have an only fans page.

Yeah.

Only fans.

Oh no.

I said no. I'm like, please don't do that. Each of the different local teams

that. Each of the different local teams has to be trying to push the envelope, but we try very hard. Again,

so there is a boundary.

Yeah. Look, ultimately we want to be a family brand. Internally, we say the

family brand. Internally, we say the words that we use is wholesome, unhinged. Wholesome just means that it

unhinged. Wholesome just means that it is like it's not overly sexualized, that it is warm, family, not not overly sexualized.

Okay?

Unhinged is that it's crazy. So, it's

like crazy but safe. We don't want to do anything that will turn people off. We

don't want it to be the case that parents tell their kids to stop doing Duolingo because Duolingo is an on fans.

I see. Yeah.

Yeah. I wanted to talk about your personal development as a CEO since it's almost like two and a half years. Any

personal development of you like you as a CEO you grow or managing the team?

Um it's funny because I just got every year I get a performance review.

You get a performance review?

I get a performance review and I just got it from the board or both the board and the people that report to me. So the executive team and the board come together and write me a performance review. Um, I can tell you

performance review. Um, I can tell you what my performance review says. Wow.

Uh, I'm I'm I'm There's some stuff that I'm good at, but I will tell you the things that I need to improve because they've been the same things for a lot of years. I'm just getting slowly

of years. I'm just getting slowly better, but not a lot. Um, probably the biggest one is that I um ultimately I'm conflict avoidant. So, I

avoid conflict. If there's something that I need to tell you that I know you're really not going to like, it's very difficult for me to tell you.

I have a very hard time telling you something you're not going to like.

And as a CEO, I sometimes have to tell you things that you're not going to like. For example, you're not doing a

like. For example, you're not doing a good job.

That is a that is a hard thing for me to to do.

That's kind of the biggest thing that I'm that I'm always being told. My

executive team has been working with me for most of them for like 10 years. So,

they know know me super well. By now,

they've learned to read me. They say

like, "Oh, there's something you don't like, but you're not telling me. So, can

you please just tell me?" Ah,

but that is kind of the biggest thing that I've that I have to get better at and I've been trying to get better at it. So I would say in the last three

it. So I would say in the last three years I've gotten better at that but I'm not yet perfect.

Is it better if you're familiar with that person can be more straight? Much better after about 5

straight? Much better after about 5 years.

After 5 years last long time I get I get good at telling you. So for example in this trip to China we just came with our chief product officer Jem. I have no problem telling him if he's doing a bad job. I'm like Jem you're not doing a

job. I'm like Jem you're not doing a good job. So, I don't have a problem,

good job. So, I don't have a problem, but it's because I've worked with him for almost 10 years. Um, but if it's a new person, it's difficult for you to It's really not good. You know, in my

time as a CEO, I've had two or three people that when I'm telling them something they don't like, they start crying.

Oh, no.

That I I'm like, it's really terrible.

I think in a lot of companies, especially when it's large, they're saying that you got to be strict with the problem you're saying and just be more straightforward. It's like a better

more straightforward. It's like a better culture. What do you think of that? And

culture. What do you think of that? And

how I completely agree with that. You

agree with that?

Yeah, I agree with that. And for me, it's not hard to be straightforward if it is about a project or about a problem, about things itself.

Yeah. But about a person, I have a hard time. So if if I think it's going to

time. So if if I think it's going to upset you about you, then I have a hard time. But if it's if it's just something like this is not working, that is much much easier. But

like you you are not working, that's very hard.

That's very hard for me.

Hard.

Yeah, it's very hard for me. Okay, I got you. Have you ever tested your MBTI

you. Have you ever tested your MBTI personality?

I have tested it, but I keep forgetting what it is. I am I am an extrovert. I am

You're an extrovert.

I am.

Oh, okay. I thought you were an introvert.

No, I am an extrovert, but I am shy, which is I am shy with new people, but I am an extin I get I get energy from people.

Okay.

And then I am also judging.

You're judging. Okay.

Um Oh, I'm more thinking than feeling.

You're more thinking than feeling.

Yes. But I'm But I'm judging.

Trying to avoid conflict with people.

Yeah. Oh, interesting. How would you describe your character? Like, do you prefer to be a product manager or a teacher or a manager or just a a tech

genius or how do you I prefer to be a micromanager?

Oh, no.

Really?

Yes. That's another thing that I have gotten better at over time. It just

turns out that after about a 100 employees, you can't micromanage everybody for sure. Yeah.

for sure. Yeah.

But if I could, I would. when the team gets larger since you like micromanaging do you feel it's more difficult to manage or I've gotten much better at it I mean

over the years I now know that in fact micromanaging in most cases is not good I now understand that but I still really want to but I've I stopped myself I also

have gotten a lot better at knowing when your executive team is really good you know that you can delegate most things so by now I know whatever problem there

is I just say you do it. It's much

better than than me doing everything.

Also, the people at least in my executive team are better than me at each of the things. For example, our marketing person is much better than me at marketing. Our product person is

at marketing. Our product person is better than me at product. Our

engineering person is better than me at engineering. So, I now know just to

engineering. So, I now know just to give it to them.

Yeah. And it's much better. But I really would like to do it myself.

Gotcha. Is it like to conquer something with the joy yourself?

I just get pleasure from solving problems. For me, happiness is solving problems. It's not the woriness about something went wrong or No, it's just it's really solving problems yourself.

Yes. And this is also another thing.

It's taking me years to learn this, but when you have a significant other like my wife, she sometimes has a a problem herself or is not feeling great, my inclination is

to sit next to her and solve the problem for her. Tell her here's what you need

for her. Tell her here's what you need to do. I have learned it's much better

to do. I have learned it's much better to just sit there and be like, tell me about it. and then and then say that's

about it. and then and then say that's really tough. Listen

really tough. Listen listen matters more and got emotional support.

Yes, that's Did you have a lot of girlfriends before that marriage?

I'm too scared to even answer that question at this point in time.

Please no.

I love you honeymoon.

Fore Fix Mr. Beast Netflix, Mr. Beast Wow.

Yeah. The handbook, the reason we wrote it was so that especially new employees and potential employees could see what it's like to work at Dualingo and what the culture is like. So the main thing

that the handbook has is five principles. We explain each of the

principles. We explain each of the principles and what's awesome about these principles is that they're not aspirational. They're actually what we

aspirational. They're actually what we do.

It's not like a principle like be the best like you know it's nothing like that. So for example one of them is take

that. So for example one of them is take the long view. Uh which is you know we want to have a company that is around for 100 years think that way as opposed to what's good for tomorrow. So that's a

big principle probably the most important one. Another one is make it

important one. Another one is make it fun. So throughout not only our product

fun. So throughout not only our product should be fun internally working at Dualingo should be fun. So that's

another important principle. Another one

is called ship it which is we really try not to sit there arguing forever about whether something's a good idea or a bad idea. They just make it and do it.

idea. They just make it and do it.

Another one is show don't tell. In many

companies the culture is such that when you're presenting something first of all you explain for a really really long time what you're doing and half of that explanation is how hard you worked. I

did this and I stayed up all night and whatever.

You don't care about that. Just deliver

the value.

Just just show me what you did instead of the whole story of how you trials and tribulations about how you got to there.

Just show me the thing. So that's

another one. And then the last one is raise the bar. Always just try to do better than we did before. So those are the operating principles and we had a whole we have a whole handbook about, you know, different examples of things we've done for each one of them.

That's kind of the culture of Dolingo.

Yeah, these these five really capture the culture pretty well. Very nice. Do

you think um as the company grows larger, it's difficult to preserve the value, preserve the culture?

Um yeah, but I think we've done a good job. I think it is still true. Certainly

job. I think it is still true. Certainly

these five things I think are are quite true.

Quite true. Yeah. Okay, that's good. I

think a lot of people may say Dualingual is good as a game, but uh to truly master a language is kind of difficult if you only spend like 5 minutes a day.

Yeah. So, a couple of things to say about that. Um, we have a lot of studies

about that. Um, we have a lot of studies that show that Dualingo is very good at teaching. You can actually learn a

teaching. You can actually learn a language and get to a very fluent level in Dualingo. So, it's very good at

in Dualingo. So, it's very good at teaching. However, there's a bit of a

teaching. However, there's a bit of a misconception. It's exactly what you

misconception. It's exactly what you said. If you only learn for 5 minutes a

said. If you only learn for 5 minutes a day, it's just going to take some time.

That is just a feature of the human brain. It takes the human brain that

brain. It takes the human brain that long to get good at a language. So, for

example, our entire French course to learn French on Dualingo, if you only do one lesson a day, which is about 3 minutes, it takes 22 years to learn it.

Uh, it works. It's just going to take 22 years. If you want to go faster, you

years. If you want to go faster, you probably want to do more than one lesson a day. So, for example, if you do two

a day. So, for example, if you do two lessons a day as opposed to one, that's only 11 years. Yes. That's like get to how I mean how what what kind of level that is a dualingo score 130 which gets

the level there is a level in which you can be good enough to communicate at work. There may be some things out there

work. There may be some things out there that tell you that you can learn a language in 3 months. That's just not true.

Yeah, that makes sense. If you want to truly master a language, you got to learn right?

You got to spend you got to spend hours.

It's just there's no if the languages are far apart. For example, English and Chinese are far Yeah.

then it takes longer. For example, if you're native in Mandarin and you need to learn Cantonese, that's much faster.

That's much faster. Similarly, if you're learning native in English and you want to learn Spanish, that's much faster.

But far apart languages, they just take longer.

Yeah, I checked my Dolingo score on Spanish.

Uh, it was only like 18, I guess, because I only do like one course a day.

That's why 18 is getting to be pretty good. Okay.

You probably now know some Spanish.

Entry level. Yeah. I can recognize some words in at the airport and I can maybe order a hamburger at McDonald's.

Yeah.

Yeah. Yeah. Yeah. See, that's it. It it

starts becoming useful and and depending on your score. So, for example, a Dualingo score, if you're at like 30, you can really reliably order things at a restaurant.

Um, if you're at, for example, 50, you can read a short story in that language.

50.

Yeah. At around 50. If you're around 90, you can watch a TV show.

I see. I'm counting like how many years I I need to get Well, you got to do more than one lesson a day. That's the thing.

a day. That's the thing.

I'm going to work harder. Yeah, for

sure.

I got to tell you, Spanish TV shows are very good. They're um they're uh there's

very good. They're um they're uh there's this thing called tel nollas in Spanish and they are uh very emotionally manipulative.

Emotion. They keep you watching because you're you're they emotionally manipulate you.

Yeah, it's like Tik Tok.

It's like Tik Tok. Yes.

I see. What are the areas Dolingo is using AI right now? um practicing

conversation which we talked about also all of our content so whenever you're learning there has to be content that you're learning it used to be made by people it is now mostly made by AI so

the content to learn math the content to learn chess all of that is is mainly made by AI uh and that means that we go a lot faster generating content so for example 2 and a half years ago from

Chinese you could not learn all the languages that you can today and we didn't even know if we were going to be able to add all these languages just from Chinese because each one just required a lot of effort.

But now we just did it much faster and it was not that expensive to add all the content to learn from Chinese because of AI.

So that's the these are the types of things that are really helping us.

Okay. How much faster? Like maybe how many courses can you made before and after?

It's hard exactly to quantify, but what I'll tell you is this. The amount of content we made in the first 12 years of Duolingo, it was all done by hand, etc. 12 years.

Uh over the last two years, we have made 50 times more content.

50 times more 50 than what we did in the first 12 years. It's a lot faster.

years. It's a lot faster.

Wow.

It just goes a lot faster.

Now I can uh maybe have a context of that because each language is like a pair.

That's right. That's the problem. The

problem is that each language is a pair.

So there's a lot of pairs. Learning

English for Korean speakers. That was a different course than learning English for Japanese speakers. And that was a different course learning English that for Spanish speakers. And each one of them we had to make now much faster with

AI and we're able to do that.

Wow. I can see that maybe you can uh expand faster in Chinese market since AI that is that is a big reason why we've expanded a lot in the Chinese market.

We're also going to be soon adding a course for Shanghai.

Uh which I am told by the people in this office that will be really popular. I

wouldn't know but the people in this office really like that.

You have no idea it's going to go popular or not?

I don't really know but they they swear it's going to be really popular.

Yesterday when I opened up the app there is a line saying that Doling will also teach endangered languages like Hawaiian. Yeah.

Hawaiian. Yeah.

Is that because of AI?

No, that is not because of AI and I'll tell you where AI it turns out that AI is not good at endangered languages.

So a lot of people have been asking us can you add more indigenous languages now that you have AI? Unfortunately AI

is not particularly good at Hawaiian because there's just not a lot of data to train. So endangered languages that

to train. So endangered languages that is not something that we've been able to speed up yet.

I see. So it's more like a social responsibility thing.

Yeah, we like we like those. So we teach a few of them. Uh we like I mean again one of the problems is there are 7,000 languages in the world. Uh there's a lot of those.

Have you ever thought in maybe in 10 years the real time translation can be quite perfect?

It's already I would say pretty much perfect.

Yeah. Then what's the point of learning a language?

Of learning a language. And what what's the point of Dolingo?

Uh so real time translation by the way is not just now. I would say for the last several years maybe about 10 years the real translation got really good about 10 years ago like Google had Google translate and it was pretty much

perfect about 10 years ago in that time we have not seen the desire to learn a language go down it has in fact gone up and I think the biggest reason is in Dualingo we have two big groups of people one big group of people are

people who are learning a language as a hobby if it's a hobby it doesn't matter I mean similar to chess you know computers have been better at chess than humans since 1997 and then there's another large group of people generally

these are people who are learning English who actually need to learn it.

So for example if you're going to be a flight attendant or if you're going to work at a hotel you need to learn English because people still need to communicate. It's not like you can wear

communicate. It's not like you can wear something that Yeah. And it's a pretty interesting

Yeah. And it's a pretty interesting thing. I mean when you're to really

thing. I mean when you're to really learn a language you know one of the ways you know you're really good at a language and I asked people for example yesterday we were at dinner somebody did not grow up in China but has been in

China for a year or two. I was asking them how good their Chinese was and they were trying to explain to me and I asked can you join a group conversation like if there's a lot of people talking about can you join a good conversation and she

said about half the time and that tells me roughly how good you are at learning language but this is for example if you're there with like a phone trying to join a group conversation it kind of doesn't work

like it just doesn't work it's awkward you know group conversations people are are saying jokes they're saying more they're adding they're interrupting each other it's very hard even with very good real-time translation.

So I think you know if you really want to be part of a culture for example if you move to China you're always going to be a major outsider if you don't know the language. Similarly I was just in

the language. Similarly I was just in this uh the CDF the China development forum I don't know Chinese and I had to listen to everything most of the CEOs would do the same.

Sure almost everybody you know the translation was done by humans and it was very fast and it was very good but you still couldn't quite get all the things and you felt Yeah. you felt like you were an outsider etc. So again, if

you actually want to, you know, move to the country or go study in the United States or something, you you actually need to learn a language.

Since you work on education a lot, uh what do you think would be the future of education because of AI? Do you think school is going to be existing or any school structural change?

So a couple of things to say about that.

First of all, you know, some people worry that well maybe we don't need to go to school because AI will do everybody's job. First of all, I think

everybody's job. First of all, I think that's many years away if that if that ever happens. And secondly, you still

ever happens. And secondly, you still need to educate people. If if you know any two-year-olds that are uneducated there, if humanity was like a bunch of two-year-olds, this would not be good,

right? It's it's chaos. So, people still

right? It's it's chaos. So, people still need to be educated in terms of learning values. They need to learn how to think.

values. They need to learn how to think.

Like for example, a lot of the early math education is just to learn how to think. Like you need to learn how to

think. Like you need to learn how to count, you need to learn how to read.

People still need to be educated even with AI. So, education will be needed.

with AI. So, education will be needed.

Schools are still needed for a couple of reasons. one, teachers are just really

reasons. one, teachers are just really good at being role models. Um, it's hard for an AI to be a role model because you're like, well, that's an AI. I don't

you don't trust it.

Yeah. You're not just not a role model.

But I think teachers are good at role models. Also, you know, during co kids

models. Also, you know, during co kids didn't go to school. A lot of times they stayed at home.

I don't think many parents like that.

Parents actually want their kids to go to school.

I think kids want to go to school as well.

It's a good thing. It's a good thing to have a place.

Yeah. to have a place where people learn how to be and parents can have. So I

just don't think schools are going to disappear even with AI. I do think that AI is going to be used more and more so that kids learn faster, kids learn better, etc. So I do think that's going to happen, but I don't think teachers or

schools are going to disappear. But

there is one other thing that is really important to mention.

Most people most people just don't like doing it. They just are not as motivated

doing it. They just are not as motivated to do it all that much. So one of the things that is really important certainly when you talk about AI is that it has to be high quality but also we have to be motivating enough that you

come back every day to do it.

Uh I think you mean the best teacher in the future is not only teaching you the course but also getting you motivated and providing some emotional value.

That's exactly right. I mean and both are needed. if it's just a really good

are needed. if it's just a really good teacher. I mean, by the way, yes, AI is

teacher. I mean, by the way, yes, AI is really awesome, but if you had a lot of motivation, the reality is that you could learn most subjects by yourself, even without a tutor.

Yes, you can just read a book.

Sure.

But most people don't do that. Like the

reality is that anybody in the world, if you have access to a public free library, you can probably learn quantum physics by yourself.

Yeah, you probably can. But most people just don't have they're just not going to sit there for 10 years doing that. So it

really matters that you can actually motivate people to do it. That is really half the battle and and so that's why we're working on that.

So you're going to replace the motivation part by gamification. That's

right. That's you're trying to do make it addicted as Tik Tok.

I would love it if we could do that because then you would be addicted to something that's positive for you. I

don't know if we're ever going to be as addictive as Tik Tok, but we'll try.

That's what we're trying to do.

I see. Yeah.

I'll tell you a few things about chess.

When somebody proposed to add chess on Dualingo was a few years ago and I said no, you should not add chess. I did not want to add chess because it's a game.

And I'm like, what? Why do you want to teach a game? I want to teach languages.

I want to teach math. I want to teach science. Why do you want to teach a

science. Why do you want to teach a game? They were disappointed and then

game? They were disappointed and then they went away as in they kept on working for the company but they you know they went to do other things.

Okay.

Um they they were disappointed. Then I

met with the minister of education of the country I was born in. I was born in Guatemala. I met with the minister of

Guatemala. I met with the minister of education. She was telling me how bad

education. She was telling me how bad the education system is in Guatemala. In

Guatemala is a poor country. The

education system is terrible. and she

said, "It's so bad that I am considering sending a chess set to every student because at least they'll learn logical thinking." And when she said that, I

thinking." And when she said that, I thought, hm, maybe we should teach chess because it actually teaches you logical thinking. And it turns out that there's

thinking. And it turns out that there's a lot of scientific studies that show that. Of course, the education system is

that. Of course, the education system is different in every country, but in the United States, the single best thing you can do to improve your grades, other than, of course, studying more. If

you're not going to do that, the single best thing you can do to increase your grades is to do chess after school. That's the single best

after school. That's the single best thing you can do to increase your grades. So, that's why we decided to add

grades. So, that's why we decided to add chess. And I went back to the guy who

chess. And I went back to the guy who had suggested chess to me and I said, I changed my mind.

What did he say?

He was very happy. He's like, "Okay, wait, let me add chess." And then I'm like, "Okay, well, you should add chess." And then he said, "One problem.

chess." And then he said, "One problem.

I don't know how to play chess.

You don't know or he he doesn't know. So, he didn't know how to play chess. He paired up with another guy who knew very little chess. No. And

the two of them combined. They're like,

"We're going to add a chess course." And

I'm like, "Okay." And they also didn't know how to program. One of them was a designer, the other one was a product manager. And they're like, "Together,

manager. And they're like, "Together, we're going to add a chess course." And

I'm like, "Okay.

They learned a little bit of chess, but not a lot. They're not very good at chess. But they use AI to create a whole

chess. But they use AI to create a whole course without an engineer. We ended up needing an engineer to put it in the app. But the course, like all the data

app. But the course, like all the data for everything, the prototypes and everything all was done with AI.

It's pretty incredible.

That's amazing. That's

Yes. And by the way, to this day, these two guys, they're not very good at chess still.

They're not very good at chess. But what

I will say is that what is good about it is that they made a course for chess that makes chess accessible to people like them.

Uh most other chess courses were made by chess grandmasters and it was too difficult.

It's difficult. It's really high stakes.

These guys made a chess course for people who are like them. You know, for example, one of the first things that if you go to the chess course, it teaches you the names of the pieces because these guys also had to learn the names of the pieces they didn't know. For

example, there's a piece that looks like a horse. It's called a knight. Oh, okay.

a horse. It's called a knight. Oh, okay.

They call it the horse. They're like,

"Well, it's the horse, but it's called a night."

Well, that makes sense. It's like you're asking an expert to learn economic terms to some layman. Then they just couldn't explain very well.

Yes. And you know what else? When we

started Duolingo, it was me and my co-founder Severign. Neither of us were

co-founder Severign. Neither of us were very good at learning languages and neither of us were like language lovers.

So, we all You are not a language lover?

No, I am not. I mean I I now have learned a lot more about languages and I have learned to learn to love languages but when we started Duolingo I was not a language lover.

How did you develop a course?

It's similar to the chess guy. Yeah. No

AI but at least so we went a lot slower but at least we made a course that was also accessible to people like us because the reality is I didn't like learning languages but Dualingo was a course that even when I didn't really

like the process of learning languages it kept me coming back. But it's because I made a course for me as opposed to somebody who is a major language lover.

They didn't need the gamification because they are intrinsically motivated. It turns out only a small

motivated. It turns out only a small fraction of the world's population is intrinsically motivated to learn anything.

Of course.

So again, this is this I think this is one of the key things about Dualingo. We

make courses for the average person and that's why we have so many users.

That's nice. That's like our goal like making videos for everybody.

Yes. And and that's the thing. I mean

there's a place for the very advanced.

They probably don't need your app. They

can learn by themselves.

That's exactly right. The people who are very who are already very advanced that or or are going to be very advanced at chess or language, they don't need our app.

Yeah. The goal is actually to keep coming back. Yes. Yeah. That's cool.

coming back. Yes. Yeah. That's cool.

That's cool.

You just went to the China Development Forum.

I did. Uh, it was very informative. It

was very encouraging. It seems like China really wants to be open to foreign investment, which is good for us because we're a foreign company. A few years ago, when we first opened the China office, nobody ever thought that China

would become the number one market. We

thought, okay, we'll we'll grow a little bit. That'll be good. But just every

bit. That'll be good. But just every year, it grows more and more and more.

And so by now, it's pretty clear that next year, China will be our number one market.

By the number one market, do you mean by uh sales or by by users? So by users, right now, US is

by users? So by users, right now, US is one and China's two. Next year, China will surpass the US by revenues. China right now is number

by revenues. China right now is number four.

Number four.

I do believe that in a few years it'll also surpass the US.

Okay. Yeah.

I saw a lot of crazy campaign went viral like for example you announced Dway is dead in the US and uh right now D is also doing a partnership with Muan the delivery app. Where are those ideas

delivery app. Where are those ideas coming from?

The local teams uh in this case the marketing team here in China is excellent. These are all local campaigns

excellent. These are all local campaigns in China in particular. It really has been one of the best countries for partnerships. It turns out that here

partnerships. It turns out that here brands really want to partner with each other and also Dualingo is seen as a very cool brand. We've just seen really excellent partnerships. We just had one

excellent partnerships. We just had one with Q Music.

Okay.

Um I think there's a a lot more coming.

I don't think I should spoil what they are, but there's a lot more coming that are pretty awesome.

That's exciting. That's exciting. Two

years ago, you mentioned uh the Dualingo English test. Any progress on that?

English test. Any progress on that?

Yeah, we're very happy with the Dualingo English test. If you don't know what it

English test. If you don't know what it is, it's a test that is accepted by 6,000 different educational programs in the world that you can take. You use it for admissions. So, historically, it's

for admissions. So, historically, it's either been TOEFL or IELTS, but now the DES, the dualing English test is accepted by the vast majority of places and it's cheaper and also you can take it online. I mean, the big progress is

it online. I mean, the big progress is that we're just accepted by a lot more institutions. And at this point, we

institutions. And at this point, we don't know this for a fact, but we believe that we have surpassed some of the the standard older tests in terms of number of tests taken. I'm very happy

with it. Uh, by the way, that's another

with it. Uh, by the way, that's another place where we use AI to prevent people from cheating.

There's a lot of cheating attempts.

We see it all the time. People are

trying to find ways of cheating and we we do a lot of sophisticated things. Um,

and don't cheat on the D.

We catch you. We'll catch you if Dolingo is not teaching languages only anymore.

Do you think people may get impression that it's less focused or Dolingo is aimed to be a super app in education?

That's the aim is to be a super app in education. We may not teach everything

education. We may not teach everything but we'll teach at least the things that take a long time to learn. I want to add science. We're not working on science

science. We're not working on science yet, but I want to add science. I want

to add drawing. You know, things that take years to learn, we want to be teaching.

I see.

I mean, generally, I want to stay mostly in education. in education.

in education. in education.

That's that's generally what what I want to do.

Do you have to approve every course?

Yeah, I feel like I should. My power to add courses is not as big as you would like as as you would think because right after languages, the course that I wanted to add was science. We added

math. After math, I wanted to add science. We added music. After music, I

science. We added music. After music, I wanted to add science. We added chess.

So, what I want is not as as important.

But yeah, I'm looking for somebody inside the company is very passionate.

Yeah. As opposed to me forcing people to do that. Gotcha. How is the course

do that. Gotcha. How is the course development process? Is somebody wants

development process? Is somebody wants to develop a course about something and do they come to you and you say yes or no or you have a committee?

Yeah. What happens is that they usually come to me and I I make a decision but they usually come with enough data to show why it makes sense. One of the most important things that I want to hear is how many people in the world want to learn this?

Uh if if it is something that only a million people in the world want to learn, I I don't want to do it. But if

it is something that hundreds of millions of people, then I'm a lot more interested. I also think it should be

interested. I also think it should be something that is good for the world.

Since you are the founder and CEO of one of the biggest education company in the world, what do you think is your quality or your trades make you successful? If

you're going to start a company and you're going to run a company and run it until it's much larger, you have to be really committed. You know, we've been

really committed. You know, we've been working on Dualingo for more than a decade. It just takes commitment every

decade. It just takes commitment every single day. One of the things that I

single day. One of the things that I think makes me successful is I'm very committed to what we're doing, kind of obsessed with what we're doing. So, I

think that helps me. I think another thing that helps me is that I am relatively good at putting in the shoes of our users. Every single day I try to learn all the different subjects. I'm

very bad at most things. For example,

I'm very bad at music and I'm trying to learn music on Dualingo. I'm not good at learning languages and I'm trying to learn languages on Dualingo. Of late,

I'm trying to learn Swedish because my wife is from Sweden. Yeah. Uh I'm not good at it, but that puts me in the shoes of the users. Gotcha.

Um because most people are not good at learning a language. So I think that is another thing that makes me good. I

would say those are probably the the the biggest to I mean there's a lot of things that I know are not good but these are the things that I know are good. Somebody said to me the the

good. Somebody said to me the the biggest thing about you is that you just work harder than everybody else and that because you're committed.

I'm very committed. I'm very committed.

Although I don't know if I work harder than the people in the China office here in Dualingo. These people work really

in Dualingo. These people work really hard.

I know they they work really hard but at least for the Western world I work really hard. Gotcha.

hard. Gotcha.

Sounds good. Sounds good. you can't

really achieve everything with yourself and you're having a great team.

What do you think your team is really good at and how did you build a team like that?

The biggest thing that I've learned about the executive team is it took me years to get there. They don't give you a headache. If I just say go fix this

a headache. If I just say go fix this problem and I never have to think about it again. I don't have to ask them

it again. I don't have to ask them whether it's been solved. They will fix it and they'll come back to me and tell me and but I just can completely forget it.

Just pure trust.

Yes. And if I tell them go fix this, I don't have to worry about it. It will

get fixed in some some amount of time.

Gotcha. In Chinese, it's called kpoo. Ku

kpoo. Ku ku. Reliable.

ku. Reliable.

Reliable. Reliable.

That's that's right. That is exactly what I the reliable. It took me years to get there with my executive team. But

eventually I have a team that is reliable.

How did you build a team like that? Any

secret sauce?

No, it just took years. I mean, you just hire people and if it's not working, you know, usually when it's not working, both parties know. So either they leave or

parties know. So either they leave or you fire them. It took some trial and error, but at this point I'm very happy with them and I want to hold on to them forever.

Forever. Yeah.

Yes.

Do you think the ability matters more or the chemistry between you guys matters more?

As I get older, I place more emphasis on chemistry and loyalty.

Loyalty than ability. Now ability should be

than ability. Now ability should be there if they can't do their job.

Yeah.

This is not good. You know, as you get older, it's not that like I'm that old, but as you get older, you start just treating things very differently. I just

at this China Development Forum, I met an investor who was much older. He's

retired. In fact,

having a meeting with somebody like that, and it's happened a few times that I meet with somebody who's either in, you know, maybe their late 70s. Those

meetings are very different than if you're having a meeting with somebody who's 30 or 40 because the person who's 30 or 40 is asking you all the tiny little details about how your company like how much profit do you make and what's your percentage of the much older

people are asking you questions like tell me about your father like that's the type of stuff they want to know and you know as you get older you start realizing that's how you really get to

know people and that ends up mattering a lot it gives you chemistry it gives you like okay I understand you a lot better because I know this is this is how it is. So, I'm starting to do stuff like

is. So, I'm starting to do stuff like that. For example, for everybody in my

that. For example, for everybody in my executive team, I kind of know how they grew up and that that tells me a lot about how they operate. For example, I have a very close and good relationship with our head of product. He's an only

child. I am an only child, too. As soon

child. I am an only child, too. As soon

as I learn he's an only child, I'm like, I understand a lot about you. Um, he's

also an immigrant to the United States.

Right there, I understand a lot about you because I'm also an immigrant to the United States. So, it just as you get

United States. So, it just as you get older, you start thinking a lot more that way as opposed to tell me your numbers, tell me your metrics. That's

starting to happen to me. Again, I'm not I'm not yet at the point where I'm 70 years old.

Sure.

But I think eventually I'll just start asking you like do you have children or not?

Oh, it really helps. It really helps because it it it it helps to see what motivates you. You know, there are people that are

you. You know, there are people that are motivated by different things. Some

people are motivated by money. Some

people are motivated by mastery, like they really want to master something.

And some people are motivated by a mission. A lot has to do with how they

mission. A lot has to do with how they grew up, you know, the way they think.

Yeah. Especially in the US, there's a very big difference between people who grew up in the US versus people who grew up in other countries. And especially if they grew up in a developing country,

there's a it's a very big difference in how they act, what they're motivated by.

And both are good. It's it's not like one is better than the other. It's just

it's good for me to understand what each person, you know, what motivates each person. That's very insightful, also

person. That's very insightful, also helpful to me as well. Yeah.

Well, yes. I mean, it really does. It's

incredible how um yeah, how much the the actual person and how they grew up. This

is why I was asking yesterday in a dinner we had here, how many people here had tiger parents.

Okay, I understand you. You had a tiger parent. You did not have a tiger parent.

parent. You did not have a tiger parent.

That's why you behave this way. That's

more understandable.

It's incredible how much each person behaves in a certain way. Whether they

have a not just like a parent with anything. What you know the relationship

anything. What you know the relationship with their parents was.

Bye bye.

Circuits being burned out.

Staying in your lane. I'm a zombie, but I still try. That's what I say.

Wait.