Abstract
Existing models of computation, such as a Turing machine (hereafter, TM), donot consider the agent involved in interpreting the outcome of the computation.We argue that a TM, or any other computation model, has no significance if itsoutput is not interpreted by some agent. Furthermore, we argue that includingthe interpreter in the model definition sheds light on some of the difficultproblems faced in computation and mathematics. We provide an analytic processframework to address this limitation. The framework can be overlaid on existingconcepts of computation to address many practical and philosophical concernssuch as the P vs NP problem. In addition, we argue that the P vs NP problem isreminiscent of existing computation model which does not account for the personthat initiates the computation and interprets the intermediate and finaloutput. We utilize the observation that deterministic computational procedureslack fundamental capacity to fully simulate their non-deterministic variant toconclude that the set NP cannot be fully contained in P. Deterministicprocedure can approximate non-deterministic variant to some degree. However,the logical implication of the fundamental differences between determinism andnon-determinism is that equivalence of the two classes is impossible toestablish.