Saturday, August 18, 2007

We are The Sims?

There was an article in the New York Times recently about the idea that we are living in a simulation, with the catchy assertion that it is actually likely we are living in a simulation. This is based on the idea that, soon-ish, we will have sufficient computing power available to simulate the brains of all the humans on Earth. From there, it follows that some of these simulations will be run, perhaps for entertainment, and, since there are more of them than there are "real" Earths, we are more likely to be a simulation than the real thing.

I can see a couple of arguments against this. First, if my life is a simulation being run for someone's entertainment, why isn't it more interesting? Either the "real" world is an astoundingly boring place, or somebody has a very unhealthy obsession with the minutae of everyday lives. Is he really watching when I pee?

Second objection: A useful simulation doesn't need to simulate every person on Earth to some arbitrary degree of realism. Especially for the purposes of entertainment, there doesn't seem to me to be any reason to throw huge amounts of computing power at simulating the day-to-day thoughts of billions of people who don't do anything all that interesting. Much simpler models will do.

Third objection: Having the raw computing power does not translate to being able to code the simulation. The task of capturing the behavior of a brain, and also simulating the reality around that brain to a level of fidelity necessary to properly stimulate the brain, is what I would call "non-trivial." As a computer programmer, my experience suggests that exponential increases in computer power do not, actually, lead to exponential increases in software sophistication. Whole brain/reality simulations on the scale that the article talks about are so difficult, I don't see it happening except for some very good reason, and I can't think of any such reason.

Those are my arguments. I don't think we're being simulated.

1 comment:

Tim Akers said...

I think your point about increases in raw computational power not necessarily being matched by sophistication of the software is pretty much the best argument I've heard against the Singularity.