The laws of physics have often been taken to be objective statements about reality, but are they? Suppose Bob is a smart computer programmer with access to a super-duper-supercomputer. In it, he simulates a world operating under different laws from that of his own. In that simulation emerges a super-duper-smart scientist Alice. Alice deduces the laws governing her simulation not realizing it is a simulation and calls them the laws of physics. What Bob calls the laws of physics is something else entirely. Without an ontological commitment to wild metaphysical speculations, Alice will have to conclude what she deduced to be the laws of physics, almost certainly have to be the laws of physics for her.
According to some ontological speculations, the shortest description which fits observational data is the actual ontological reality. Certainly a law like Schmidhuber's computer program which multitasks over all possible programs, one of which simulates our universe, will have a shorter description than a program which only describes the laws of physics of our universe? By this "kolmogorov" ontology, wouldn't Schmidhuber's program be more likely to be ontologically real as the laws of physics? If there is some measure over observers, it would also means the laws of physics (not the ultimate) will have to be typically generic while giving rise to sentient observers. What about a quantum computer running a quantum superposition of all possible programs, with eventual decoherence between the different programs? That would also have a shorter description.