I'm 19, a first-year CS student, and I have a fully homomorphic encryption dApp live on-chain.

I also took my first "What is Machine Learning?" class today.
People assume you learn in order. School, then internship, then real projects. Foundations first, hard things later.
That's not how it happened for me.
I started in Web3 at 14. By the time I reached university, I'd already closed deals, shipped products, and learned to read a smart contract before I'd sat a single formal CS exam.
This year I built PayParity: confidential salary benchmarking on Zama's FHE stack. It lets you prove you're underpaid without revealing your salary. Live and verified on Sepolia.
It took 11 failed deployments to get there. Type mismatches. Forgotten access controls. A compare op that blew past the gas limit. The docs don't warn you about most of it you just fail, read the revert, and try again.
And today, after all that, I sat in an intro class learning what machine learning even is.
My skill tree has holes you could drive a truck through. I'm filling them out of order, in public, where everyone can see the gaps.
I used to think that was a weakness. Now I think it's the whole point. You don't need permission or a perfect foundation to build. You need to start, ship, fail, and fill in the rest as you go.
What did you build before you felt "ready"?