You’ve been there. You sit down to study, open your textbook, read two lines, and your phone buzzes. Suddenly, 45 minutes have vanished, and you’re watching a video about how to build a swimming pool in the jungle. You feel guilty, but you’ll do it again tomorrow.
For years, ed-tech companies have tried to solve this by building a better digital textbook. They tell you to be disciplined. They tell you to turn off your phone.
Traditional ed-tech assumes learning is the priority and games are the temptation. Gizmo decided this is completely delusional.
Meet Gizmo, an AI study app built by 7 Cambridge graduates that just racked up 13 million users. They didn’t build a better teacher. They built a better game engine that happens to teach you.
The secret isn’t a breakthrough in AI pedagogy. It starts with removing friction. If you have 50 pages of notes, making flashcards takes hours. You burn out before you even start memorizing.
Gizmo takes your notes, PDFs, YouTube videos, or even your existing Quizlet decks, feeds them to the AI, and instantly generates a gamified quiz.
But AI generation alone doesn’t get you 13 million users. The real magic happens next.
Gizmo didn’t make the learning tool look more like school. They made it look like a mobile game.
You start answering questions. Get it right? Keep going. Get it wrong? You lose a life. You get 15 lives a day. Run out? Wait for them to regenerate, or pay for more. There are streaks, XP, leaderboards, and multiplayer battles with your friends.
Suddenly, reviewing your biology notes feels exactly like climbing the ranks in a competitive mobile game.
They aren’t competing with Anki or Quizlet. They are competing with TikTok, YouTube, and every app fighting for a student’s attention.
You might think this is just a Duolingo clone. But there’s a massive difference. Duolingo produces its own courses. They decide what lesson one is. That makes it incredibly hard to scale to your specific college organic chemistry class or your friend’s CPA exam.
Gizmo doesn’t produce content. The users bring it. The AI converts any content into the exact same game format. Today you play history; tomorrow, anatomy; the day after, traffic laws for your driver’s license.
A 60-year-old woman used it to memorize bird species. Starbucks even used it to test baristas on coffee facts. The AI made it possible to apply proven game mechanics to virtually any domain for near-zero cost.
This growth wasn’t fueled by massive ad spends. The 7-person team went to where students already were: TikTok. They made content about study hacks, not hard product pitches. Then they weaponized network effects. One student makes a deck, shares it with a desk-mate, they start competing on the leaderboard, and the entire study group gets pulled in.
One by one, 300,000 users turned into 13 million.
Retention in ed-tech is notoriously brutal. Students download an app before finals and delete it right after. Gizmo solved this by making the act of returning to study the core game loop. Your streak can’t break. Your ranking is under threat. Your friend is challenging you.
They aren’t trying to reduce young people’s screen time. They are trying to turn it into better screen time.
This is the ultimate blueprint for product-market fit in the AI era. Stop trying to invent a behavior that doesn’t exist. Find something millions of people already do but hate doing, use AI to delete the hardest step, and wrap it in consumer psychology that makes it addictive.
AI didn’t just give Gizmo a fancy feature. It changed their cost structure. They don’t need to hire 100 teachers to cover 100 subjects. They just need to perfect the converter that turns any content into a playable experience.
The next time you design a product, ask yourself: are you lecturing your users, or are you giving them a high score to chase?
FAQ
Q: Doesn't gamifying education just create empty dopamine hits instead of real learning?
A: It's a valid fear, but Gizmo relies on proven cognitive science mechanics like active recall and spaced repetition. The game mechanics don't replace the learning; they just provide the fuel to make you actually do the boring, effortful part of studying.
Q: What's the practical takeaway for building an AI product today?
A: Stop trying to invent new behaviors. Find a tedious task millions already do (like making flashcards), use AI to eliminate the friction, and use consumer psychology to make them come back. AI changes your cost structure, letting a 7-person team scale to 13 million users.
Q: If Gizmo is just Duolingo for everything, why hasn't Duolingo crushed them?
A: Duolingo is bottlenecked by content production—they have to build every course themselves. Gizmo's users bring their own content (PDFs, notes, videos), and the AI instantly converts it into a game. It's a totally different scalability model.