True random number generation is a fancy term with little real world use. It has to do with a new event having nothing to do with a previous events. This fundamentally clashes with the ability to control it; if the new event ignores everything you’ve done to set it up, how could you expect it to happen in a way you could detect it?
The power of computers is their predictability. A 3 GHz processor has something like 3 billion calculations happening each second. If even .0001% of those were wrong, that’s 3 thousand wrong numbers every second. Modern computers are expensive because of the effort to eliminate these errors as much as possible.
These combine into computers to being bad at true random number generation, even if it is only a little worse than the rest of our reality (a coin flip is dependent of the physics of the coin launching, traveling through the air, and landing as an example of not being true random). However, us humans are even worse than computers. We perceive a pattern, and we expect it to repeat. This is why true random number generation has so little real world use; even if humans saw true random numbers, we wouldn’t know it to see it. So the “pseudo random numbers” that computer programmers have figured out are good enough for most of us.
Latest Answers