To have a simulation all we need is an emulated brain AI and another AI that recreates the world the emulated brain sees to make the emulated brain believe what it sees is real. This adversarial AIs will be run many times while training.
By the time we manage to simulate a world and everyone on it this kind of single person simulations will have run so many times that the chances that we live in a world simulation or in base reality are minuscule by comparison.
I’m no longer concerned that we are playing full dive vr since that seems more difficult to accomplish than this scenario, and therefore the chances are much smaller. Although I think the chances of it are still higher than world simulation or base reality.