Many a scientist has hypothesized that we are living in a computer simulation (especially Nick Bostrom who says it’s almost definite.) If a computer were simulating the rules of physics, our molecules would be virtualized bits of data. There’d be no way to tell. It’s hard to wrap our minds around what this reality would really entail, because we don’t know enough about the nature of our universe’s physics, or about what kind of computing is possible in the future. I would argue that we are not capable of comprehending both of these things fully.
But even with our tinny tiny understanding of the nature of reality, the concept reveals an interesting take on the nature of data. If we can be represented by a complex model, then in a sense we are the complex model, encoded in matter. Data means any form of organization whatsoever, and what we call matter is the medium of storage of the data. It could be bits and bytes, it could be an idea in somebody’s head, it could be a chaotic and unorganized mesh.
But what of the medium that encodes the data? No matter what you put on a hard disk, it is still a hard disk. If it is simulating a universe, doing math, or booting up and then sitting idle, its fabric and basic structure is the same. And — and this is what interests me now — it exists in a world where entropy is steadily decreasing. We believe that the world always tends towards less entropy, less organization. A pile of sand will eventually scatter, a concentrated drop of liquid will eventually disperse, unless external energy is used to keep it together (the physics definition of entropy is more exacting, but it’s a phenomenon that works on any level.) So, too, the “entropy” of a hard drive is always decreasing – it requires electricity and repair to remain in the same clean baseline state.
SO…if we have a computer simulating our universe, it takes a significant amount of energy just to keep the laws of physics going, irrespective of what is happening in the universe. This entropy could be tied in some way to the physics of our own universe.
The alternative I see is if scientists were able to construct a computer whose very structure did the computation. One example is a non-mechanized coin sorter – the rules of the system are built into its structure and it does not require a continual amount of energy just to continue to enforce them. I think this is an important distinction to make between different kinds of computing systems, and we may very well begin to see more advanced structural computers come about one day.
This analysis will seem woefully simplistic to future generations – probably even our own. I saw Fassbinder’s “World on a Wire” at MoMA last year, which was a result of deep and serious thinking but, because it was made in 1973 before most computers, now looks almost idiotic. Who knows? I just like to think about things in taxicabs.