If the tokens keep getting faster but the Rust compiler stays slow I am probably going to ditch Rust for everything that is not mission critical Bullish for Rust inner loop into python bindings (what we use for the amm challenge)

I have also become increasingly frustrated with hosting providers that have slow deployment times. I want to let the agent try to debug in a loop and often this requires making multiple deployments. The debugging loop can turn into 80%+ waiting around