 Severin, we're back on stage together. We've become stage buddies over the past years. Somehow it's a duo in finale, you and me, sitting at tech conferences and speaking about Duolingo Europe and what's next. You said there was a special way you wanted to start today's talk, so over to you. Yes, actually I have a question. So question for the audience. If you are a Duolingo user, show me with your hands how long is your streak. So one means 10 days or longer, two means 20 days or longer, 10 means 100 days or longer. Wow, that's a lot of long streaks. Anyone above 1,000? Yeah, there, awesome. Nice, nice, nice. I have a surprise for you in that context, Severin. So the last time we were on stage together, we were at another European tech conference. And I asked Severin, how did you build a brand that everyone in the world knows? And he said, it's the owl. And I was, maybe I was nervous. I froze and I really just didn't remember that the owl is the icon, is the mascot of Duolingo. And so I blacked out and I said, which owl? And Severin was a bit stunned on stage. I think he was a bit cross with me at the time. And so what I did this time is I went to the gift shop of Duolingo on their website. And I posted the largest order ever for stuffed animal Duolingo owls. And I knew that he would start the talk by asking for the streak. And so right now in the audience, those with a 1,000-day streak, raise your hand because you're going to get a Duolingo owl because we had a box coming into Helsinki. And here's one for you. You get to give it to anyone. Or maybe, actually, I have someone that should get it. And you might want to give it back when you hear that story. Because this morning at the hotel reception where I'm staying here in Helsinki, I asked for that package. And the receptionist handed over the package. And she said, I have to ask. There's little Duolingo owls on that package. What is this? And I told her the story. And I told her that I was going to be here. Of course, she knew about slush. She's from Finland. She works here in a hotel. And she said, how cool. You're on stage with the founder of Duolingo. You know what? I have a streak of 1,483 days. And that was for me once again a reminder of how remarkable it is what you've built, Severin. And I quickly introduced myself. My name is Alex. And the reason I'm here is because I run a venture capital firm that invests only in Switzerland. And Severin happens to be from Switzerland. And when we started out, we asked all Swiss entrepreneurs to support us in boosting the small ecosystem. Severin was so kind to do that from the first day on. And what you've built in the meantime is incredible, Severin. You are the CTO co-founder, member of the board, of a stock listed company that is the largest language learning application in the world, most used, most loved. And what I find amazing about it is that with a lot of tech things, we tend to have love-hate relationships. Social media, we're worried about our mental health, Google, too many ads. There's always some negative to it. But everyone loves Duolingo. Do you pinch yourself sometimes that this is your reality? Sometimes, yeah. It is truly remarkable. I mean, I obviously did not expect it to be ever-dissuccessful. And what's remarkable is, truly, everyone loves Duolingo. And it's a good company for the world. I truly believe that. And that's why I'm still with the company after IPO and all that. And your IPOed, I think your IPOed at US $6 billion at NASDAQ. And there was a bit of skepticism, if I may say that so bluntly. Obviously, everyone was wondering, monetization, is this going to be a real business, or is it just an app, a productivity app? And I checked the market cap this morning, and we're beyond US $9 billion. And your IPO was a year and a half or two years ago. So in the roughest of times where most tech stock was punished severely, you managed to increase your market cap by 50%. That is insane. Like, what's going on? What's happening? Why is the world so, so keen on Duolingo today? Yeah, it's a good question. And by the way, we don't look at our stock price every day. And it's not a KPI that we try to optimize. What we try to do is achieve our mission, which is to provide the best education and make it universally accessible. And going back to why we're a public company, but we have an operating principle that is take the long view. So we want to build a long-lasting, hopefully 100-year-lasting company. And today's stock price is irrelevant if you take the long view. But why it probably went up is because the market rewards execution and the profitability and all of that. So that is good. But yeah, we've been just executing really well since the beginning, I would say. Well, congratulations. And thank you that you're here with me today. We're in an audience of students, founders, young entrepreneurs at the beginning of their journey. There's also some tech decision makers and investors here. So I thought we'd speak about three things in the next 20 minutes. And I would love to go a bit back into your journey, starting the company. How was that building Duolingo? I want to speak a bit about where does Duolingo stand today? And what are you working on? What are the product roadmap topics that you're discussing, business model changes, a bit out of the boardroom? What's going on at Duolingo? And then the last third of our talk, we need to talk about something that in 2023 you cannot ignore, which is AI and how that relates to what you're doing at Duolingo. So let's get started with the look back. You studied in Switzerland at the Technical University in Zurich, and then you moved to the US and you started your business there. Why did you choose to go to the US and understand you co-founded with your professor? How did that come to happen? Can you tell us the founding story, your path from Switzerland to incorporating Duolingo and also maybe a time stamp? When was that? Yes, so I've always wanted to study in the US. So I grew up in Switzerland, small town, close to Zurich. I've always wanted to study in the US for a little bit. But I went to ETH in Zurich, which is a very good technical school there. And it was good. And they didn't have an exchange program with the university I wanted to go to in the US. But the sister university of ETH had that exchange program. And I managed to get myself into that exchange program. And that exchange program is really what led to everything else down the road. So it was an opportunity of a lifetime. Obviously, I didn't know that at the time, but I've always wanted to study there, study in the US. And that exchange program was between EPFL and Carnegie Mellon University. And at Carnegie Mellon University, I met my eventual co-founder, Luis. He was one of the youngest professors at the time at Carnegie Mellon. And when I met him, I introduced myself as I'm Severin Hacker, and he still remembers that. And that's how I met him. And he was the smartest person I had ever met up until then. And I was like, if I can ever work with this guy, I'll try. I will do that. And he convinced me to stay at Carnegie Mellon for grad school. So he was a professor. I was his grad student. And two years into my PhD, we started working on what eventually turned into a dual and go. Nice. And so your professor is your co-founder. And my understanding is he's been at that time already a serial entrepreneur that had a very successful exit to Google. So this is someone that has a really high standing within the startup community in the US, probably made it easy to raise first funding and get some attention to dual and go at that time. But how was the dynamic between the two of you as co-founders? Because it's not the typical story of three students that as equal co-founders start something. But there's been a bit of a hierarchy difference between the two of you. And how has that been at the beginning? And how has that evolved until today? Because your two co-founders and both of you are still in the board, even though the company is now publicly listed, which is not often the case. Yeah, great question. And by the way, most of the time this doesn't work. So usually when there's a professor and a grad student starting a company, the professor owns 99% and the grad student owns 1%. And the grad student does all the work. And it never works. It never works. But in our partnership, it was equal from day one. And we had a little contract, actually, that I still have at home, where we wrote down how we make decisions at dual and go, how do we do it? Who is responsible for fundraising? Who is responsible for this and that? And we wrote this down. We signed it. It's not a legal contract. You would see from a lawyer. But it worked. And it's magical. It probably prevented tons of arguments and just having a clear division of labor and role and ownership. That's, I think, the secret to this relationship. This is interesting. And actually, in fact, I didn't know that. We never spoke about this before. But when we invest as a venture firm, one of the things that is extremely important to us is an equal share split between co-founders. Because the moment it is not equal, any discussion turns very emotional, because there's some sort of power dynamic that is not necessarily healthy. So that's cool to hear the confirmation of that, even in a setup like this. Now, Duolingo IPO today, you raised some funding in between. If you look back at the journey, and you're speaking in front of entrepreneurs here, what are the big milestones, the ups and downs, the memorable moments that you think were key to becoming the success story that you became today? So I only remember the ups. Yeah. No. So obviously, first one was funding of Duolingo. When was that? 2012. 2012. OK, nice. So 2012. And we started as a website. And it was growing slowly. It was nothing magical. But it was a pretty good product, I would say. But it was a website. So next milestone was mobile. And I remember the day before we launched our first iOS app, Luis and I said, oh, what do you expect from this launch? How many users would we get? And Luis and I agreed on maybe 5% of our user base will be on mobile, 95% will be on web. And within, I think, a week or two weeks, it was all mobile. Crazy. And in that year, we grew 14x in one year. So mobile was a huge milestone for us. And fundraising must have been a piece of cake that year. Yes. Because we raised on $0 revenue, which is, by the way, better than a little bit of revenue, at least at a time. Things have changed. But at a time, no revenue was better than a little bit of revenue. And then we raised on the user growth. Next milestone was subscriptions. So figuring out monetization. When was that, roughly? I think that was 2016. 16, yeah. So four years into the business, you started to generate revenue, or you took the decision we would try to generate revenue from that user base. Yes, yes. And it was a very hard decision internally. Because as I said before, we were a mission-based company. And we hire all of these people that are motivated by the mission. And when we said we're going to have ads, they were like, wait, we can't show ads. That's not the mission, right? We shouldn't make money, you know? And so how did you motivate internally? Did you tell people that it's a means to an end, that you need the revenue, that you need the business model behind it in order to be able to reap growing and executing on that mission? Or how did you sell that? Yes, so we said the mission is to provide the best education, make it universal accessible. That doesn't mean that there are no payers. What it means, it means that we don't put content behind a paywall. So that anyone who wants or needs to learn a language can do so with Duolingo, whether they have a bank account or credit card or not. So, and this is still true today. So we obviously monetize, and we're, you know, monetization is doing super well at Duolingo. But you don't have to pay for content. So the way we monetize is by a feature-based subscription. Yeah. So subscription was the big... That was the big game changer. It was a big, yeah. And did that work immediately? Because I guess a lot of entrepreneurs would be really scared. You've been showing a track record of having users, of getting people excited to use your product. But then, you know, are people ready to pay money for it? Is that a moment where you have a few sleepless nights and a bit of nervosity today? Or it worked pretty quickly, actually. So, and we weren't the first subscriptions. So they were already, you know, yes, body-fied, et cetera. Things were, people were already used to subscriptions to a degree. But for us, it was mostly an internal challenge. Once we built it, it was, you know, it took off very quickly. And did you use that traction to then raise funds again? Yes. So it's inflection points. First, you went to mobile, massive growth. You raised funding. Then you launched subscription revenue from nothing too relevant. You raised funding. And then you raised, probably, last round pre-IPO. Is that more or less the funding story? Yeah. And what's funny is, because for the first five years, we made no revenue. People just assumed this is going to be, this is a research project by two academics. And they're going to sell it to Google, because my friend herself thought the past two companies to Google. So I was like, I think there was even a tech-run story at some point. You know, they're going to sell to Google or whatever. But the moment we made significant revenue, we realized we don't actually have to sell. And we can stay independent and go to the IPO right. And follow that mission that you had defined when you started, Luis and you. And that's interesting, because we've gotten to know each other better and better over the past years. And we've had some conversations where I was just out of curiosity asked some of the VC questions you would ask someone like yourself. It's like, did you have funds that passed on you? And now in retrospect, you're like, ha, ha. Look at what we've done. And the list of those funds, there was a lot of big brand names that they just didn't believe that you guys are going to turn this into a business. If you were a founder, and you're sitting in the audience right now, and maybe you're raising your seat round, or you say, would you tell founders, don't worry too much about which VC is backing you. Get the money, stick to your mission, and keep executing. Or is it really important to convince those top tier funds that what you're doing deserves their attention and money? Later. So we are, I don't like to wear strong believers that you should hire the best people. Only work with the best people. Not just for employees, but investor. Work with the best investors. Work with the best law firms. Only work with the best, always. Interesting. And so that was then, was that discouraging? But obviously, stay true to the mission. Yeah, exactly. And sometimes it's tough, right? Because when investors sit in front of you and say, you're not generating revenue, we're not going to invest in you. And your mission is to not focus on revenue, but focus on the impact you can create with people trying to learn language. OK, Duolingo today, last time we were in stage together, it was roughly six months ago, and your PR department told me I'm not allowed to ask anything outside of language. And I was a good boy, and I stuck to it. But I understand now we're allowed to talk about it. And there's math and music available to learn via Duolingo. How did that happen? Yes, so as before, our mission is not just about language learning. It's all of education. And for the first 10 years, we focused on language learning. But we always wanted to expand to other subjects. And finally, this last year, we got to a stage where we felt comfortable adding new subjects. And the way we did it is the Duolingo way. So we added those subjects to the main learning app. So you don't need to download another app or install it, whatever. It's the main learning app. And you just, instead of learning Spanish, you go into the core switcher and say, I want to learn math, or I want to learn music. And all the game mechanics we've built for Duolingo, they're there too. So your streak transfers, your gems and XP, all of that transfers, the leaderboards, et cetera. So you have all of these engagement features that have been so powerful for Duolingo available for these new subjects as well. And is there a bit of a hybrid to that, if I may ask critically, in the sense that language learning is one thing. It's a lot about, I think, memorizing, hearing how someone says it. Now you could argue math. There's also an aspect of memorizing. But there's also a deeper understanding of the rule set, maybe similar to grammar. But still it feels different. Music feels different. It almost feels a little arrogant to say, you know what? Now we're Duolingo. We're going to teach everything now. It's easy as pie. Have you tested this for a long time? How have you built that product? How do you avoid that you defocus, that you derail attention of your team? Now I understand it's within the same app, but internally, suddenly you're not a language company anymore. Suddenly you're a learning company. And that is a change of focus. How do you look at those questions? Yeah, so we start with, by the way, right now it's just math and music. And we start with those two because we believe the Duolingo method of how we teach languages can be applied to those two other subjects. So the Duolingo method is very much based on implicit learning. So learning from examples, learning by doing. And this works well for math, for example. This is not all of math, but it works for the arithmetic, fractions, multiplications, et cetera, et cetera. So it's a basic math we believe can be learned efficiently with an adapted version of the Duolingo method. But this way you drift off a little bit into children as focus users. Is that a correct assumption? Because obviously during school time, I don't know if anyone here raised your hand if you're really into learning math. It's probably a lot less than those that want to learn Spanish. So is that also a conscious choice to go into guiding students during school years through learning process and helping them? Yeah, it's a good question. And math has a bad brand, has a brand problem. So a lot of people have bad memories from math and school, et cetera. And that's the one thing they definitely don't want to do or learn. But people, if it's one of those fundamental skills that anyone on the planet at some point needs to learn. So yes, we're still trying to figure out, what is the math product for adults and what is the math product for children? There could be a math product for VCs, because every year we have some crazy hype where a lot of VCs forget about math. It's my understanding. And in the past years, we had last mile delivery and metaverse. And I'm really glad NFT trading. I'm glad we're out of this bubble. But this year, it's AI. And everyone is talking about it. Everyone wants to invest in it. Everyone is also worried about it. And the first headline we had for this talk here was Duolingo disrupted by AI, question mark. And then Severance PR departments said we cannot do this. Has to be more positive. And we said we're still going to openly speak about this. Is it something that you're worried about? And we see chat GPT going and translating better than Google Translate does. We see them scripting whole episodes of The Simpsons with generative AI. So are you worried that language learning is going to be changed? Hopefully it will be changed. And I mean, I'm looking forward to this world actually. I think it's a great opportunity for Duolingo. AI is a huge opportunity for Duolingo. And let me talk to two points about this. One is we want to teach with technology. So if you want to achieve our mission, we have to rely on technology. The best way, by the way, to learn almost anything is with a one-on-one tutor. And it's very expensive. This is how the kings in the early days, in the 19th century, this is how they taught their kids. They hired the best teacher. And it was one-on-one tutoring. And to this day, this is still the best way to learn almost anything. In the world. Prom, it's very expensive. So what we want to do with Duolingo is use technology and AI to replicate that efficacy of a one-on-one tutor inside Duolingo. So basically, not going to disrupt you. Not going to replace you. It's just one more tool in the toolkit of the Duolingo engineers and product team to think about how they can create that personal tutor for every single user of Duolingo. And then the other part of AI is, do you still need to learn a language? And is that why you're going to math and music? Are you worried that people are going to stop learning languages? When I was young, people said you need to learn French in order to work in France. I think nowadays, English to work anywhere in Europe is going to be sufficient. Do you see a trend there? Or is the interest, the hunger of people to learn other languages to fit into other cultures? Is that something that you think will not be broken by good automated translation? So that was not the motivation. And there's two things, and it was a previous point as well, is one thing that's interesting about Duolingo, our users, especially in the United States, is if you ask them, what would you do if Duolingo went away? That went away. Which other app would you use? Guess what they say? Oof. Bubble? No. No, none of them. None of the competitors. They say I would spend more time on Instagram, TikTok, social media. So we're competing. We're not competing with the other language learning apps. We compete with people's attention, for people's attention. And yeah, so we compete with social media. And so why do they learn a language? Because they have fun learning a language. They have fun learning a language. And that is remarkable. And I see the timer blinking, because that means we're out of time. But I think it's a beautiful finish, because what you've achieved deserves so much gratitude. You've turned something that is really, really hard. I mean, all of us here in the audience, I don't know about you, but learning vocabulary for our French test, it was painful. And Duolingo turns this into fun. And even the receptionist at the hotel this morning told me that she's on a modern 1,000 streak. Like I said earlier, everyone loves Duolingo. And it's become a competitor not to other learning apps. It's become a competitor to watching videos, playing games, or being on social media. And that's amazing. Congratulations, everyone. Thank you for being here on stage with me today. And thank you all for listening to us. Thank you.