Abstract
The notion that all (or in weaker sense, some) natural phenomena can be modelled as a computable process, some kind of algorithm is recently gaining scientific recognition, and more research is dedicated to the rigorous explorations of the mapping between natural phenomena and the formalised computational systems. There is some debate and controversy as to how much of the natural can be expressed in the models of the artificial, although due to formalised nature of mathematics and physics itself, it is generally accepted that computation is viable way to model physical reality. Contemporary developments in computer science and in physics not only do no refute computationalism – they provide more data and evidence in support of the basic theses. In this article we discuss some of the aspects of contemporary computationalist efforts based on the traditional notions of Turning Machine computation. Then we present an extended notion of computation, that goes beyond the traditional Turing limit. We propose a new interactive computation model called Evolvable Virtual Machines (EVMs). The EVM model uses the notion of many independently asynchronously executing processes, that communicate between each other and with the outside environment. We present some of the pitfalls of traditional computationalism, and compare it to our new, extended computationalist model, based on the notion of massively concurrent interactive computation (hypercomputation). We argue, that hypercomputationalism based on the collection of asynchronously concurrently communicating computational machines is a more compact and more appropriate way of representing natural phenomena (or the Universe in general). It is theoretically sound, and does not violate any of the current state-of-the-art physical theories. We discuss the details of our computational architecture, and present some of the implications of the hypercomputationalism on contemporary physical, life sciences, and computer science.