Modern C++ is one such example.
It has a track record of adopting paradigms that are then rejected by consensus (strong OO).
It adopts modern practices because other people like them. Not because it has practical evidence that they are worthwhile.
That statement is 60 years old and its truth continues to outlive its utterer: https://en.wikipedia.org/wiki/James_E._Thornton
I suspect that once the day comes when the hardware folk are no longer regularly giving us software folk a free lunch, there will be a rediscovery of the importance of "knowing what one is doing". But I fear, not before.