5 ms·

I wonder if the right vocabulary is to talk about local optima, even though that isn't really the problem with being incrementalist as a rule. Or maybe once upo

by rz2k 11y ago

I wonder if the right vocabulary is to talk about local optima, even though that isn't really the problem with being incrementalist as a rule. Or maybe once upon a time "paradigm shift", also not the right concept. Anyway, my point being that constant iteration or the use of heuristics have value, but it can also be useful to understand models, and occasionally take a break to think big about the direction you want to move in the long term, with the realization that many of the steps to get there will violate some pet heuristics.

gozo 11y ago
You can't really have a paradigm shift without a foundation. Apple had a long history in software before the iPhone, SpaceX and Tesla have been leveraging the US long history of car and space industry etc. As far as I know the US does not have a very good foundation for public infrastructure projects nor the cooperation between public and private entities. What you need to do is, like China, buy the best proven technology currently available. Then after you've made that work, take all that knowledge and invent the next big thing.