I was kinda skeptical as well, because I actually don't speak much Chinese (even though I learned it at a young age), so I think in English, yet I was very good

I was kinda skeptical as well, because I actually don't speak much Chinese (even though I learned it at a young age), so I think in English, yet I was very good at math in HS (competitive). Yet, looking back, I realized that I did think of numbers like "four two" instead of "forty two", so doing calculations like "[four | two] plus [two | eight] equals [six | ten] -> 70". Hence, faster. It’s actually interesting — I kind of see this article as more of an analogy. I'm don't think it's the main cause of a performance gap, but it definitely could contribute. Imagine: what if this article was about developing a typical business app with C# vs. C++ — instead of “you can count from 0-10 faster”, it’s “you can manipulate strings faster” (measured in dev time, not CPU cycles). Developers will switch to newer languages that make it more efficient. Eg, Paul Graham's arguments about how LISP made Viaweb successful. So, I'm curious. Developers will switch languages because they're more efficient and expressive. But it's hard to switch languages. Not impossible, though. Will English evolve to become more efficient over time? Or will we all switch to a newer language because it's more efficient?