Most “AI slop” isn’t a model problem. It’s a feedback problem.
I made a video on automated verification loops in Claude Code.
You'll learn how to make the agent check its own work and fix it before you ever see it.
Link below 👇
Do you want to become a Solana Developer?
I'm thinking of scheduling a livestream watch party of my Solana Crash Course, spread out over 1 week.
I'll be there live to answer any questions you have along the way.
And I'll add in additional lessons based on live feedback.
My goal is to get you ready to ship apps in 1 week.
Anyone interested?
will be posting daily updates of projects Im building on insta, so if you want to learn and build along with me before long form tutorials come out, go follow :)
Uniswap-style AMMs lose traders money on stablecoin swaps. Not from fees — from math.
x*y=k spreads liquidity across every price from zero to infinity. But USDC/USDT trades at $1 — so almost all of that liquidity sits where no trade will ever happen, and traders eat the difference as slippage.
Curve's StableSwap invariant fixes this: flat near the peg (nearly zero slippage), constant product at the extremes (pool can't be drained).
New whiteboard video: the math visualized side by side, then building a working stable swap AMM on Solana with Rust and Anchor.
Full video linked in this thread 🧵👇
What actually happens after you submit a transaction on Solana?
Well I broke it down for you.
In just 1 minute you'll be able to understand exactly how Solana transactions work ↓
if you've ever wired microservices together, you already understand how Solana apps work
you just don't know it yet
new whiteboard video on CPIs — how programs call programs onchain 👇
Ever wonder how an app's data is stored without a database or a sever?
On Solana, data is stored directly on the blockchain with Accounts and PDAs.
Learn what a PDA is in the video below, or save it on YouTube to watch later 👇