Simulation Argument
Over at 'she sells sanctuary' Gianna muses about metaphysics and refers to the 'Simulation Argument'. (Gianna refers to a News Ltd story on the matter but I can refer you to the original paper by Nick Bostrom on the topic). Gianna refers to the fact that a simulation could be running somewhere in our 'multiverse' (referred to in the News Ltd article), which is a nice way for astrophysicists to push there pet theories, however theories of the multiverse are are completely unneccessary for the argument.
In a strange twist of fate I was thinking about this argument only last week.
The argument runs, basically;
IF
1. You accept that it is possible for a consciousness to be simulated on some sort of computing device independent of the brain (this is called substrate independence) and that the sort of computing power to perform this task many times over is possible to obtain in this Universe (or some other Universe).
AND
2. You accept that some species of post-human life will exist in the future, or alternatively within the bounds of the univers, and that it will be desirable for some species of post-human life to run consciousness simulations.
THEREFORE
3. The probability that you yourself are running in such a simulation is close enough to 1.0 (i.e. certainty) that it is almost impossible to admit otherwise.
There is an equation, and an argument, in the original paper which shows this point, which I will not reproduce here (go to the paper to read it).
Of course if 1 and 2 are correct in metaphysical reality, and you don't believe it, it doesn't make it untrue, nor if you believe such things as 1 and 2 that they are correct. It's just that if you hold that 1 and 2 are true, then you should logically believe that you are in fact simulated.
Actually the paper argues about the factors in point 2 more than point 1. That is the author's equation states that one of the following must be true:
- the number of civilisations reach a post-human stage of development ~= 0 (ie humans are to become extinct)
- the number of civilisations who do reach such a stage and are interested in ancestor-simulations ~= 0
- the probability of you being inside a simulation ~= 1 (ie an almost certainty)
My interpretation of teh argument, although not supported by the author's original conclusions whcih deal only with the points above, is that really the Simulation Argument is about the computability and substrate independence of human or animal consciousness. It has nothing to do with Multiverses, although I would hazard a guess that if the Multiverse theory was proven (how I don't know) it would increase the probability of Simulation even closer to 1.0.
The only way I see any person escaping the belief that they are in fact a simulated human is that either you believe;
- humans are doomed to extinction without a successor species, or at least incapable of developing an advanced technological civilisation (pretty pessimistic, but given current human behaviour, a possibility).
- human successor species will not be interested in their antecedents (I think a distinct improbability), or will be prevented from doing so by moral strictures (as it inflicts suffering on the simulated awarenesses)
- the simulations are imperfect by some necessity and a simulated consciousness can tell the difference (ie detect it is a simulation) (this is pretty interesting, but how do we gain this information?)
- available computing power is much more finite than we think (I don't think this is likely)
- consciousness is not substrate independent (goes against the grain of most mainstream thought in this field).
Anyway it's an interesting little conundrum and funnily enough, opens the possibility for a scientific humanist empiricist to reintroduce the concept of the Creator - Programmer as Creator. The programmer in me likes that even as the philosopher revolts!