swestrup | Hunting the Elusive Glorkum

As we progress in technological prowess, many new ideas are tried first as simulations before being tested in the real world. In order to make these simulations more applicable to the real world, great effort is expended to make them more realistic. As more radical ideas need to be tested, larger and more complex simulations are undertaken.

So far, so good. There isn't a lot contraversial there. But, if one takes this line of reasoning to its logical conclusion, you end up assuming that at some point we'll have simulations that are so sophisticated that we can simulate a human mind. If we invent the mind from scratch, that's what futurologists call Artifical Sentience (to distinguish it from the Artificial Intelligence, which was supposed to deliver just this, and never has). If one instead copies an existing human mind using some sort of scanning technology, this is what the futurologists call 'Mind Uploading'.

In either case, the simulated environment might be good enough that, without being explicitly informed, the simulated mind might never know its world wasn't real. This idea shouldn't be too unfamiliar to anyone who has seen "The Matrix", "eXistenZ" or even "Total Recall". According to some peoples calculations, we should be able to simulate a human mind on a $1000 computer by the year 2020. Right now computing power is doubling roughly every 12 months. It'll probably be doubling faster by then, but we can take the cowards way out and not try to predict that rate and just stick with the 1-year doubling. So, after just a bit more than 33 additional doublings in power after we can simulate a human brain, we'll be able to simulate 10 billion human at once (on a desktop computer!)

So, by about the year 2050, if not before, we should be able to simulate the estimated population of the Earth for that year. Now, $1000 computers are pretty common nowadays, and by then may well be ubiquitous, but lets be conservative. Right now there are about 600 million $1000 computers worldwide, and an estimated 13 million copies of "The Sims" (a runaway hit game that simulates a bunch of humans) have been sold.

If we assume that our 'Planet Sim' also has an installed base of 13 million units. This would mean that for every 'real' person living on the planet at that time, there would be 13 million simulated people. If you chose a random inhabitant of Earth, the odds would be 13 million to one against that person being real. This is why some folks have proclaimed that, looking at the odds, it seems far more likely that we are currently living in a simulation than not.

Looked at another way, if that randomly chosen person is you or me, then there is an equally good chance that we aren't real. We would never know. To a being in a simulation, the simulation would seem completely real. Even if their environment differed from how reality worked, unless they had a baseline to compare to, they would never notice. Remember that in "The Matrix" deja-vu is an artifact of the environment that the 'real world' doesn't have. No one in the matrix takes this as evidence for the existance of the matrix. Its just accepted as a normal part of the environment, just as we accept it in our world.

Does that mean that there is no way to tell if one is living in a simulation? In theory, that is correct. But, as scientists are won't to say "In theory, theory and practice are identical. In practice, they differ." Every non-trivial simulation I have ever seen (and the vast majority of the trivial ones as well) have had bugs in the simulation software. So, whereas a perfect simulation might make it impossible to know if you were in a simulation, an imperfect simulation might have a bug that leaked the information.

Now, it couldn't be just any bug. Remember that "deja-vu" in the matrix was arguably a bug caused by poor editing of reality. So, even though their may be glitches caused by bugs in the simulation, we couldn't go looking for them as we wouldn't know that we had found one. No, the only way a being in a simulation could tell that they were in one would be to find a fundamental inconsitency in reality. A place where the rules broke down completely.

Now, I do have a name for such an inconsistency, and that name is 'glorkum'. There is a very old and extremely complex computer game called Nethack, which is essentially a simulation of a fantasy dungeon crawl. Being large and complex, this game has had many bugs over the years, and shows no signs of running out of them any time soon. One thing the authors put into the game was an object, "the glorkum" that you should never see. The game never generates a glorkum on purpose but particularly nasty bugs can create them. A glorkum appears whenever the simulator tries to manipulate an object, but discovers on inspection that it has no idea what it is, or what its properties are.

Now, in Nethack they created glorkums so that folks looking for bugs could find them and squash them, and so that the computer wouldn't try to manipulate an object whose type was corrupt. If our reality is, in fact, some sort of simulation, and if the creators of the simulation aren't perfect, there may be glorkums or their equivalents hanging around somewhere. I'm not sure what one would look like in our universe, but I tend to imagine a scintillating point in space-time where the laws of physics fail to hold.

It is conceivable that a determined group of scientists might even be able to figure out just what a glorkum might be like in our universe, and go looking for them. Not finding one wouldn't prove anything. I've played nethack for years and never seen one, although I know for a fact they exist. On the other hand, finding a glorkum would prove that we were in a simulation and could have profound effects on our worldview. That said, I have grave doubts that hunting for glorkums would be a good idea. After all, everyone that I know of who found one in Nethack usually discovered that finding one caused the simulation to become completely corrupt and to crash shortly thereafter.

So, I have no plans to go looking for glorkums, even if I knew what to look for and how to search, but I have to wonder if they are out there, somewhere, waiting to destroy the universe if we ever stumble across one.