You’ve watched the explainer videos. You’ve read the Wikipedia article. You’ve nodded along to the animations. And yet — you still don’t really get it.
That’s exactly where I found myself with the Kakeya conjecture, one of the deepest open problems in mathematics and a recurring theme in Fields Medal research. The videos were beautiful. The diagrams were elegant. But the connection between Kakeya sets and harmonic analysis? It slid off my brain like water off a duck.
The gap between understanding a concept and feeling a concept is where most education dies.
So I did what any AI enthusiast would do: I asked Fable 5 to do the research and build me a game. Not a textbook. Not another video. A game I could play with my hands and feel in my intuition.
The result is a deceptively simple interactive experience at quantisan.github.io/kakeya-game. You manipulate sets in a plane, trying to cover every direction while keeping the area as small as possible. That’s the Kakeya conjecture in its bones — how small can you make a set that still contains a line segment in every possible direction?
Here’s the twist that hit me while playing it: the game itself is a Kakeya set. Think about it. A minimal interface — a few controls, a canvas, a score — that somehow covers all the directions of understanding you need. No 40-page paper. No 30-minute lecture. Just a puzzle that makes your hands do what theorems describe.
The best educational tools don’t explain complexity — they make you inhabit it.
This is the part that should make educators, AI researchers, and anyone who cares about democratizing knowledge sit up straight. We’ve been treating AI as a summarization engine, a code generator, a chatbot. But what Fable 5 did here is something different: it translated an abstract, Fields Medal-tier mathematical concept into a tangible, manipulable object. It built a bridge between the symbolic and the sensory.
Traditional education says: read this, then read this, then maybe in three years you’ll understand. The AI-game paradigm says: play this for ten minutes and feel the conjecture in your fingertips.
I’m not claiming you’ll solve the Kakeya conjecture after playing. I’m claiming you’ll understand what the conjecture even IS — and that’s more than most graduate students can say after a semester of harmonic analysis.
We don’t have a knowledge problem. We have an interface problem. And AI just handed us a new one.
The implications go beyond math. Every intimidating field — quantum mechanics, information theory, advanced cryptography — has a game hiding inside it waiting to be extracted. The question is whether we’ll use AI to build more PDFs or whether we’ll use it to build experiences that make people feel smart instead of stupid.
I know where I stand. Play the game. Then tell me you didn’t learn more in ten minutes than you did from every explainer video combined.
FAQ
Q: Isn't a game just a gimmick? You can't really learn deep math from clicking around.
A: You're confusing 'learning' with 'mastering.' Nobody claims ten minutes of play replaces a PhD. But the game gives you the one thing textbooks and videos consistently fail to deliver: an intuitive grip on what the problem even is. You can't build on a foundation you can't feel.
Q: What's the practical takeaway here? Why should I care about a niche math game?
A: The takeaway isn't the Kakeya conjecture. It's the method. If AI can turn a Fields Medal problem into a playable interface, it can do the same for any complex domain — your company's architecture, your product's user flows, your research's core tension. The format is the insight.
Q: Isn't this just AI hype? A clever demo that doesn't scale to real education?
A: Every paradigm starts as a demo that doesn't scale. The first website was a demo. The first app was a demo. The point isn't that this specific game replaces education — it's that it reveals a format nobody was systematically pursuing. The people who dismiss demos are usually the ones who get disrupted by them.