> abstractions ultimately leak This is where they really loose me. No, they don't. Only bad abstractions leak. How did real limitations with statistics ("all

> abstractions ultimately leak This is where they really loose me. No, they don't. Only bad abstractions leak. How did real limitations with statistics ("all models are wrong, some are useful") mutate into pessimism about designed, not discovered, systems? When did we all become such mindless empiricists?