Because game engines are for games. They fake a lot of stuff in their models.
The brilliance of humans is their ability to turn airports into Excel spreadsheets.
This bit of the thread is interesting - "The increasingly blurry boundary between games and life".
I'd say elections are going a bit this way, along with a lot of other stuff.
A 'monoculture' would be bad, but I don't think that's a real risk. Also, Direct3D11 isn't going away.
We've been surprised to see how many uses people have found for the tech, wildlife research, commercial drone operations, and more.
The default is currently cesium, but if you click the "view beta player", button, you'll see our hand-rolled game engine.
It is very much constrained to geo type applications so the "eating the world" is a little dramatic. I like the hong kong airport example that...shows how broadly applicable geo type stuff can be and how much applicability it has if you take it to a granular level
It used to be that game engines had real-time processing requirements (60/30/15 frames per second) and were constrained by the hardware on which they ran, but movies looked more photo-realistic and could take hours or days to produce a single frame.
As silicon and algorithms get faster and closer at approximating reality, the difference between the output of the two softwares approach one another. Given the amount of effort required to build these softwares it makes sense to save the years of effort, cost involved and risk involved with what essentially amounts to building a bespoke, internal graphics / physics / particle simulation engine.
