Toward a Model of Models: Part I

In this work we discuss the issue of what, if any, limits exist on modelling. We extend notions of Godel to suggest that there may exist theoretical limits to what can be modelled, independent of the complexity of the modelling system. Second, in this talk we briefly describe the system called VPMS, currently under construction by our group, for performing modelling in a distributed HPC context. We suggest that the theoretical limits to modelling may be overcome by appealing to learning based computing paradigms in non-Turing complete systems.