> But Bloomberg Intelligence estimates it would take about eight years to move just 10% of Apple’s production capacity out of China, where roughly 98% of the company’s iPhones have been made.
98% of the iPhones that have been made already, not 98% of iPhones currently being made. So the percentage is probably somewhat lower as of present.
And because of the "separate sites" approach to mobile web development. Good for some media-focused applications, but usually the separate mobile site has an oversimplified interface that can't be zoomed in or out.
Arguably the neural net is the leakiest of all abstractions. Inputting enough data, preventing overfitting. Not having a human readable implementation because of the computer figured it out with brute force. Those are all complex details leaking out of the abstraction. Machine learning has its uses in processing large amounts of data, but it is the opposite of a designed algorithm with self-contained abstractions.
Surprised the article doesn't mention Isaac Asimov's Foundation in which psychohistorians try to predict the future with mathematics, similar to Turchin's idea.
Seemingly random studies in biology can have unexpected benefits. Studying gecko toes led to the invention of a super-adhesive. https://geckskin.umass.edu/
Local maximum instead of global maximum. In order to become more fit, it would have to temporarily decrease its fitness and thus is stuck in a dead end. More mutation or larger population size in that specific simulation might solve the issue.