The prototype for the chess course was so engaging and fun that it convinced us to build it into the app.

With AI, we can generate high-quality educational content quickly, allowing us to expand into new subjects like chess, math, and music, enhancing our learning offerings. We started working on the chess course eight months ago, initiated by two people who had no coding or chess experience, and they managed to prototype it successfully before we added engineers to productionize it.

Learning to code enhances logical thinking and problem-solving skills, making it beneficial even if AI can automate coding tasks in the future.

We were able to create the chess course in just nine months, the fastest time ever for a new course at Duolingo.

In the early days of Duolingo, we realized that keeping users motivated was the biggest challenge in language learning, which led us to incorporate game mechanics into our lessons.

We have a lot of data on how people learn on Duolingo, and while we use it to build models, we can probably do more with it to enhance AI progress.

AI will help us overcome cognitive biases, allowing for better decision-making and a more accurate understanding of reality.

Games serve as an aesthetic form of thinking and doing, enabling players to pursue goals, solve problems, and understand systems in an engaging way.

At Duolingo, we have a rule that it's better to have a hole in the organization than to hire someone who is not a good fit, which has helped us maintain a positive culture.