Building Fermat's Library was a side project that we quickly developed over a weekend, aiming to get it out there as fast as possible.

The goal of Fermat's Library is to promote open science, which includes sharing data and making research more accessible to facilitate replication and collaboration.

I never aimed for profitability; I aimed for efficiency. We kept our headcount flat and focused on creating a fun work environment, which allowed us to move quickly and effectively.

We aimed to be profitable on the first order, which was uncommon at the time, and this focus on profitability helped us succeed.

Our goal is to democratize education, and while we initially thought we could monetize through user-generated translations, we found that a freemium model was more effective for our growth.

We spent the first five or six years of Duolingo not monetizing, which I regret; we should have started monetizing after year three to be further ahead today.

We encourage contributions to Fermat's Library by inviting users to annotate papers, spread the word, and create lists of papers for their research groups.

Fundraising has never been a goal for us; it's just a tool to get the resources needed to build the product and service we want.

At that time, there wasn't a clear monetization route for Reddit; we didn't have advertising or any real way to generate revenue, so we operated as a lean startup to survive with scarce resources.