The overall purpose of this blog is to investigate the question "Is the universe a computer?" or rather a little bit more scientifically phrased, "Can we understand and describe the universe as some kind of computational process is some as yet unknown, possibly discrete, substratum?
I will call this The Computer Science Paradigm in Fundamental Physics.
This is also a book writing project I've been working on for a couple of years. It started in earnest in 2004 when I gave a semi-popular course at Skidmore College about the many connections between theoretical physics and theoretical computer science. I got the idea to edit the scattered lecture notes into something readable. Be that as it may, as I hope will become clear in the first few postings here, speculating about the universe being a computer is not very interesting. What is interesting from a scientific point of view is to formulate a critique of that set of ideas.
Research, discussions and speculations along these lines go back at least three decades. There are lots of scientific papers and books on the subject. These days, the speculation is so ubiquitous that people at dinner parties no longer find it strange if one brings it up, but just nod their heads and turn to another everyday subject. Indeed, the speculation is well known through popular articles and in particular through science fiction films and novels. The reason for this is of course the ubiquitousness of computers themselves and the historical fact that dominating technology tends to color the world view. The computer as a metaphor, as it were. The ever more sophisticated computer games with their virtual reality adds to the familiarity with computer generated worlds.
Now, ubiquitous or not, the claim that "the universe is some kind of computational process in some as yet unknown discrete substratum" contain at least two caveats. Take the last one. We do not know what that discrete substratum is, or would be, we don't even know if it is discrete, although there are indications in contemporary fundamental physics that it might be. The first caveat, having to do with computational processes, is the fact that the theory of computation, whether classical or quantum, severely restricts what can be computed even in principle, and the question arises as to whether all physical action falls within the computable realm or not. We don't know. The "universe as a computer" idea rakes up quite a lot of issues in fundamental science, as well as things we simply don't know. That is one reason why a critique is interesting.
So if one wants to be scientifically honest, it is way to early to claim the universe to be a computer. But the idea is interesting and it is certainly intriguing. It is well worth investigating using the tools we have available; experimental knowledge of fundamental physics, theoretical physics, mathematics, the theory of human knowledge, natural philosophy and last but not least, computer science itself.
It seems to me that many articles delving on this subject actually fail to take computer science seriously enough to really apply it forcefully to the question. Hopefully, one thing that is new in my approach is an attempt to do just that. Furthermore, before briefly looking ahead, let me stress that I'm not claiming that the Universe is a computer. That is not at all the purpose of these notes. There are certainly quite a few texts that do just that, more or less seriously, i.e. puts forth a particular "theory" of a computational universe. While many of these texts are perfectly honest and serious, the speculative nature of the idea attracts also not-so-serious authors, and there is spectrum of texts moving into the outright wacky. Instead, the purpose of my investigation is critical. It is not to put forth yet another speculation.
Looking ahead
There are three circumstances that, at least in a general way, lends support to the Computer Science Paradigm (I will refer to it as CSP for short). These are often not discussed explicitly, and to get going, I will just bring up one here, and introduce the other two in later postings.
The superficial way in which we describe computational processes and physics systems is almost the same. We are using what could be described as an input-transformation-output model. The descriptions are almost the same because of the following two dichotomies, which are related to the caveats just mentioned.
- The continuous - discrete dichotomy
- The classical - quantum dichotomy
Physics, whether classical or quantum, use the mathematical continuum in its models of physical systems, whereas classical computer science is wholly discrete. There is thus a mismatch of basic tools and concepts. Classical computer programs can only approximate models of physical systems.
Some authors cut this Gordian knot by simply declaring that classical physics and classical computational theory is simply wrong, and that we should use quantum physics and quantum computational ideas throughout. Quantum theory use the mathematical continuum but describe properties of discrete, quantized entities. Indeed, using quantum theory, physical systems and quantum computers, seem to be almost indistinguishable. This might very well be a correct point of view, but before buying into that option, I would like to conduct the investigation in the pages ahead.
The second dichotomy in particular, touches on the fact that all human knowledge, when it is explicitly formulated, must be recorded using some language, in practice using a combination of either formal, mathematical or natural. But language is classical even when what is described is quantum. This leads to a conundrum that is closely related to questions of interpretation of quantum mechanics. (Niels Bohr stressed that classical language is unavoidable in any description of the results of measurements on quantum systems.)
So in order to investigate whether the paradigm is scientifically sound, we must understand how, and if, these gaps can be bridged (if indeed they need to be bridged and not just be discarded as mere fictions of human thought). That is one of the objects of the following notes. In order to do that we must understand present day fundamental physics, and try to see a little bit into the future where it might go. We must also understand what computer science is about. Most likely, both computer science and fundamental physics will develop in the near future. Whether their paths will converge or diverge, we do not know.
There is one source of hope, for those who prefers to think in those terms, in the realization that computer science is really the science of data and processes in general. Any data, any process. Thus it might eventually be powerful enough to encompass everything, even fundamental physics. Or seen the other way around, this last statement might be simply tautological. We simply have to do more research.
