Jonathan Blow

I covered this in my talk ... there is no evidence of such efficiency, in fact, it is very much the opposite.

which talk pls

I once criticized languages 50x+ slower than Rust/C. Got lots of negative feedback. People seem to really like their slow languages. Running the whole internet (servers and clients) isn't insignificant source of energy consumption. We could definitely do better.

I feel we should also account for developer productivity, though. If it takes 50x the developers to make the same thing in C; how much resources did it cost to have those extra developers? How long would a CPU need to run a 50x slower program to match that? Quite long, I bet.

I am not willing to accept that good productivity must equal bad runtime performance. We can surely design lanauge with both good productivity and good runtime performance. Rust for example showed that added safety didn't kill performance. It did the opposite (proves noalias).

That is completely fair, hoping for that language to appear one day too! C, however, is not it, wouldn't you agree?

C is not it. However, this idea that newer languages have made productivity much higher, falls into the category of religion and not science. Everyone believes it but it has not been measured and if you start looking, productivity of modern programmers in these languages is