How to Make Learning as Addictive as Social Media | Duolingo's Luis Von Ahn | TED

 I'm from Guatemala, not Guantanamo—different place. Guatemala’s below Mexico, a smaller, poorer country with better Mexican food. Education often widens inequality: the rich buy elite schooling, perpetuating wealth, while the poor barely learn to read, staying poor. I got a rich person’s education despite not being rich, thanks to my single mother’s sacrifices, leading me to a US college and a PhD in computer science.

Ten years ago, as a Carnegie Mellon professor, I started working with my PhD student Severin to give everyone equal access to education. We chose to teach foreign languages first, not math or computer science, because two billion people globally learn languages (80% English), and knowing English can directly boost income, unlike math, which requires further steps. For example, a waiter who learns English can work at a hotel and earn more.

To reach everyone, we built Duolingo, a mobile app, since smartphones are widespread and building schools is too costly. Duolingo’s freemium model lets anyone learn for free with ads, while wealthier users in places like the US or Canada pay to remove ads, effectively subsidizing education for poorer users in countries like Brazil or Vietnam—a small wealth redistribution.

Smartphones, however, compete with addictive apps like TikTok or Instagram. To make education engaging, Duolingo uses their psychological tricks, like streaks (counting consecutive days of use) to keep users returning, with over three million users maintaining streaks over a year. Japan has the longest streaks; Latin America, the shortest. Notifications, timed 24 hours after last use, remind users to learn, and a passive-aggressive message after seven inactive days (“reminders aren’t working, we’ll stop”) brings users back.

Our green owl mascot, tied to these notifications, has sparked memes and SNL skits, embedding Duolingo in pop culture. While educational apps can’t match TikTok’s addictiveness, they don’t need to—learning’s inherent meaning compensates, making 80-90% engagement enough. Duolingo has more US language learners than all US high schools combined, a trend true globally.

I hope we can apply this to other subjects like math or physics, especially those learned through repetition (e.g., reading, elementary math). Gamifying repetitive learning works, while explanatory subjects may need high-quality videos, like Sal Khan’s work. My vision is a future where screen time delivers high-quality education to all, rich or poor, via smartphones. But please, do your language lessons today.

Q&A Response
Subjects learned through repetition, like reading or elementary math, can be gamified like Duolingo. Most meaningful skills require thousands of repetitions, making them ideal for engaging, repetitive app-based learning. Explanatory subjects are harder but can use high-quality videos, as Sal Khan does.