Runtime code generation and execution in Go: Part 1 https://t.co/T6BF6Q5QUM
ok just published the very first part of my blog series on the weird world of runtime code gen/exec in pure Go. Check it out and hope you enjoy it!
LLMs are insanely good at minimizing large fuzz test cases at the scale and speed we have never seen before. The way they are doing it not deterministically but semantically correctly is exactly what I was doing manually by hand when working on wazero's compilers
so frustrating that i am like at 99% ready of my super cool build tool to open source, yet i have spent the last few weeks on that last 1% (codex/claude maxed out and 6 machines running 24/7). I believe this is going to be a game changer but without that 1%, it's not proven yet
Visited Grand Canyon for the first time and the view was so breathtaking that it gave me a few ideas on the spot and ended up talking to Claude right up there on the edge of the Canyon lmao
@KohaviYuval that's because I am building something that needs extremely large cpu intensive test corpus and need to run multiple benchmarks to iterate. Usually I run agents only on two of them and others are for just running tests
so frustrating that i am like at 99% ready of my super cool build tool to open source, yet i have spent the last few weeks on that last 1% (codex/claude maxed out and 6 machines running 24/7). I believe this is going to be a game changer but without that 1%, it's not proven yet
I am super excited about the fact that no one gives a shit about Wasm Component model now. One of the textbook level disasters in the tech history and so glad that no one living in a real world didn't take it seriously
I don't think I know a single person who owns a Tesla and drives anymore. They use full self driving 90%+ of the time.
Incredible to think that our kids are going to grow up in a future where all they will know is autonomous driving