Towards a Computational Monism

The thought occurs to me that Max Tegmark's Computational Universe Hypothesis and Daniel Dennett's computationalist model of the human mind can serve as a basis for a clarified ontological/epistemological standpoint.

Since Democritus, philosophers have wondered about the basis of the world, whether it come down to elementary particles flitting about in the void (Democritus, Epicurus, etc.) or whether there be concepts needing to be added to this perspective like "Prime Movers", etc. (Aristotle, Aquinas, etc.). The word "dualism" has many meanings. There is a particular extremist form of Cartesian dualism I won't get into here. For the purposes of this post, I will use the word "dualist" to refer to any "atoms + something else" viewpoint (to simplify) and monist to refer to a "just elementary components" viewpoint. This over-simplifies, but what I want to do here is to suggest a new monist position to counteract the various dualist, "Prime Mover" type positions out there.

Simply put, the computational universe hypothesis of Tegmark suggests that all computational universes exist. Which is to say, any and all universe models that could be modeled computationally (on a Turing machine, to be technical about it) exist. This is superficially similar to Platonism, except that it says these computational worlds are ALL THAT EXISTS, there is no "Platonic cave" or other dualist notion. We simply live in one of an infinite number of model worlds that can be described as Turing Complete. Turing Machines replace the Democritus/Epicurus "void" in this picture and the bits of information thereon replace the "atoms" of Democritus.

Since by definition one Turing Machine can run any other Turning Machine upon it, the world becomes recursive. The universe we dwell in is (in a sense) a computer program, but so are we (see Daniel Dennett) also computer programs (Turning Machines). And, as some of us are ourselves programmers, we have programs being written by programs within a program. (Where the word "program" loosely means "Turning Machine" - one would have to write a book to be more technical about these matters and I haven't the time or the inclination at the moment, ha!)

This is a "monist" position because EVERYTHING from minds to atoms to the void to everything in between is describable as a Turing machine (possibly, in the case of the Universe, an infinite one). Thus we need not have ephemeral notions like "souls" or "ectoplasms" to account for our minds, since there is no real "objective vs. subjective" distinction here anymore. It is one big recursive Turing Machine, or an infinite set of these recursive machines. Now to play "devil's advocate" here, I suppose one could posit a "Prime Mover" type entity as one of these Turing Machines that are "out there" in the Tegmark computational multiverse. But I will say that while one can certainly do that, it is not an "information additive" operation to make, that is, such a Prime Mover would have nothing to "move", be it the universe or our thoughts. It would be "part of the system", and not something "added" to the "system" itself. So, while one could posit such an entity, it would be an operation such that the "additive value" would be lost so that "Prime Mover" would have as much usefulness as "angels on the head of a pin". But I had to address that issue, since, if philosophy since Democritus has taught us anything, it is that those "Prime Mover" advocates don't go down easily. :-)

There is also a "sneakier" consequence to this view. I have "sneaked time in the back door" by saying the universe could be one big Turing Machine since (by definition) Turing Machines operate "in time" - they are not "timeless". "Time" for a Turning Machine is the operations it makes, and "space" is that upon which it makes its operations (the punch card, whatever). The copious consequences to this means that "in the real world" time is an actual physical entity which can be (as in anti-De Sitter spacetime) distinguished from space. Which is not to say that other timeless models cannot be used (Euclidean spacetime for instance) but that these are models and not the "real world" which always, being a Turing Machine (in this view) has a well defined notion of "time" (if not a well defined reference frame). This applies to us too. It means that the "soul" or the "personality" is energia and not ousia in Aristotle's terms. We are "the computing" not "that which is computed". This might sound like philosophical gibberish with little practical import, but just look at all those AI enthusiasts signing up for cryonics (including Marvin Minksy for goodness sakes). If we are the data, cyronics might work. But if we are the operations upon the data (energia rather than the data itself, ousia) then cryonics could not work even in principle. So in this view the folks who hope to live forever via cryonics are better switching their bets to supporting quantum immortality, not perhaps a likely scenario, but still more likely than cyronics since it is not forbade in principle, which cyronics would be if something like the picture I am proposing is correct.

Well, all of this is probably deserving of deeper explication and perhaps I will later but hey, the title is "Towards a computational monism", not "A definitive guide to computational monism". Oh, and if someone already has this term used for something else, then I could call it, say, "Computational Neurocosmic Monism". How about that?

This intro is admittedly brief, but then again it took me 20 minutes to write and the first three decades of my life to arrive at, so there you go. :-) But, fanciful as it may seem, it is about as good a monist position as I can currently construct, and, if one will forgive a bit of audacity, it is I feel at least as good if not better as the various Prime Mover-inspired positions that have arisen from time to time.

In "Scream" the villain says to Neve Campbell's character Sidney Prescott, "It's all one big movie, Sid". I might say, "It's all one big Turing Machine, Sid." :-)