typical DNNs work like function approximators. Obviously they can approximate the trajectory of planets even with a simple architecture. It is certainly an achievement to have those computations successfully performed using discrete fields, which may be more in line with quantum theory. But I don't understand, why that leads to a prove of simulation theory.
typical DNNs work like function approximators. Obviously they can approximate the trajectory of planets even with a simple architecture. It is certainly an achievement to have those computations successfully performed using discrete fields, which may be more in line with quantum theory. But I don't understand, why that leads to a prove of simulation theory.