In this paper, I investigate a prefix-free universal Turing machine (UTM) running multiple programs in parallel according to a scheduler. I found that if, over the course of the computation, the scheduler adjusts the work done on programs so as to maximize the entropy in the calculation of the halting probability Omega, the system will follow many laws analogous to the laws of physics. As the scheduler maximizes entropy, the result relies on algorithmic thermodynamics which connects the halting

more »

... probability of a prefix-free UTM to the Gibb's ensemble of statistical physics (which also maximizes entropy). My goal with this paper is to show specifically, that special relativity and general relativity can be derived from algorithmic thermodynamics under a certain choice of thermodynamic observables applied to the halting probability.