Similarly, “We can’t even model the inside of a proton. How can you possibly know when a pot will boil?”

Turns out you can get pretty far modeling the macro behavior without perfect knowledge of the micro details.