How to Make Learning as Addictive as Social Media | Duolingo's Luis Von Ahn | TED
my personal experiment
Greetings, everyone! I want to take a moment to introduce myself and share a little about my background before diving into a subject that is close to my heart. My name is Luis von Ahn, and I hail from Guatemala. For those unfamiliar, Guatemala is a country located just south of Mexico. Now, I often find myself having to clarify something amusing yet important: Guatemala and Guantanamo are two entirely different places. While Guatemala is a beautiful country full of rich culture and history, Guantanamo is where a U.S. military prison is located—far away in Cuba. These two are often confused, but I assure you, they’re not the same.
Let me give you some context about Guatemala. For the Americans in the audience, you might find this analogy helpful: Guatemala is to Mexico what Mexico is to the United States. In other words, just as the United States deals with immigration concerns regarding Mexico, Mexico has similar concerns about Guatemala. The dynamics are strikingly parallel. Guatemala is a smaller, poorer country, but it has one thing going for it that I can proudly boast about: we have even better Mexican food than Mexico itself.
Despite its rich culture, Guatemala struggles with systemic poverty, and this brings me to an observation about education. Around the world, people often view education as the great equalizer—an instrument to bridge the gap between the rich and the poor. However, I see it somewhat differently. In practice, education often widens the gap. Here’s why: individuals with wealth can afford top-tier educational opportunities, perpetuating their economic advantages. Meanwhile, those from underprivileged backgrounds often receive inadequate schooling, barely learning basic literacy skills. Consequently, their prospects for upward mobility remain limited, particularly in developing nations like Guatemala.
I count myself as one of the lucky ones. Although I didn’t grow up wealthy, I had the privilege of receiving a high-quality education—something usually reserved for the elite. Being an only child, my mother, a single parent, poured every resource she had into my schooling. This sacrifice enabled me to attend college in the United States and ultimately earn a Ph.D. in computer science. This transformative experience deeply influenced my worldview and planted the seeds for a life mission: providing equal access to education for everyone, regardless of their economic background.
This vision materialized about a decade ago when I was a professor of computer science at Carnegie Mellon University. Alongside my Ph.D. student, Severin Hacker, I decided to tackle this enormous challenge. We started with a simple question: How do we give everyone the opportunity to learn? Education, as a whole, felt overwhelmingly broad, so we decided to narrow our focus. We thought: What should we teach first? Initially, we considered math, as both Severin and I have a strong passion for the subject. However, we quickly pivoted.
Ultimately, we chose to start with foreign languages, and the reasons for this decision might surprise you. First, there is a massive global demand for language learning. Around two billion people worldwide are learning a foreign language, whether through formal education or informal means. Of these learners, approximately 80 percent are focused on English. This makes sense because, in many countries, proficiency in English can significantly enhance one’s income potential. For instance, a waiter who learns English may transition to working in a high-end hotel and earn more. Unlike math or physics, where mastery often requires years of study before translating into economic gains, learning a language can yield immediate financial benefits.
Another crucial factor in our decision was accessibility. Building physical schools worldwide would be prohibitively expensive, but smartphones offered a scalable solution. Today, a large proportion of the global population owns or has access to a smartphone, and this number is steadily growing. Thus, we decided to create a mobile-based platform that could teach languages to anyone, anywhere, at little or no cost. And that’s how Duolingo was born.
To ensure Duolingo was truly accessible, we adopted a freemium model. Users can learn as much as they want for free, but they might see ads at the end of lessons. For those who dislike ads, there’s an option to subscribe and turn them off. Interestingly, most of our revenue comes from subscribers in wealthy countries like the United States and Canada. Meanwhile, users in poorer countries, such as Brazil, Vietnam, and Guatemala, generally stick to the free version. This model creates a subtle yet meaningful form of wealth redistribution: well-off individuals essentially subsidize the education of those who cannot afford to pay.
While this approach makes education accessible, we encountered a major challenge: smartphones are inherently distracting. They come preloaded with some of the most addictive content humanity has ever created—TikTok, Instagram, and mobile games, to name a few. Competing with such captivating distractions is like serving broccoli alongside the most indulgent dessert. To succeed, we realized Duolingo had to make learning feel as rewarding as entertainment.
We achieved this by borrowing techniques from the very platforms we were competing against. For instance, we implemented streaks, which track the number of consecutive days a user engages with the app. The psychology here is simple yet effective: people are motivated to maintain their streaks because losing them resets the counter to zero, which feels like a personal failure. Today, millions of Duolingo users boast streaks that exceed a year.
Another tactic we employ is notifications. While these can often be spammy and irritating, in the context of education, reminders can be genuinely helpful. Our AI-driven system carefully selects the optimal time to send notifications and crafts personalized messages to encourage users to return. Interestingly, the most effective time to send a notification is 24 hours after a user’s last session—essentially the same time they were free the previous day.
In cases where users stop engaging altogether, we send a final notification that essentially says, “We’ll stop reminding you now since it doesn’t seem to be working.” This passive-aggressive approach, funny as it sounds, often prompts users to return, feeling as though our mascot—the green owl—has given up on them.
Speaking of the owl, its personality has become a cultural phenomenon. Users have created countless memes, and even SNL has featured skits about the owl’s relentless pursuit of language learners. This cultural resonance is a testament to how we’ve managed to make education not just accessible but also engaging.
Looking to the future, my hope is that we can replicate Duolingo’s success across other subjects, from math to physics. Repetition-based learning, in particular, lends itself well to gamification. While we may never make educational apps as addictive as TikTok or Instagram, that’s okay. The inherent meaning and value derived from learning provide additional motivation, bridging the engagement gap.
Today, Duolingo has achieved something remarkable: in the United States alone, more people are learning languages on our app than in all high schools combined. This is proof that accessible, high-quality education delivered via smartphones can transform lives. My ultimate dream is a world where screen time becomes synonymous with meaningful learning, allowing every individual—regardless of socioeconomic status—to access the education they deserve.
So, if you take one thing away from this talk, let it be this: education is not just a privilege; it’s a right. And before you go, please, pretty please, don’t forget to do your language lessons today



Comments
There are no comments for this story
Be the first to respond and start the conversation.