Topic: Varia
Simulation: noun the representation of the behavior or characteristics of one system through the use of another system, especially a computer program designed for the purpose.
Today I would like to present to you the argument of a certain Nick Bostrom about whether or not we live in a simulation. For those of you who are not aware of this philosophical question, it goes a little like this; how can we be certain, as thinking beings, that we are not in fact living in reality, but in a simulated reality part of a larger world, that we may in fact simply be living in a computer program. If that were the case, then there would be no way of knowing. So which is it more logical to think: that we live in reality, or that we live in a simulation? Dr. Bostrom makes a point of explaining that it is more logical to think the latter, and here is his argument.
Of the following three prospects, only one can be true:
(1) Sufficiently advanced civilizations which can create simulations do not, and cannot, exist;
(2) Sufficiently advanced civilizations which can create simulations would not be interested in doing so (i.e. they would not create simulations);
(3) We are most likely living in a simulation;
Now, before we get started, we need to clarify what it means to be a "sufficiently advanced civilization". To be one such civilization is to have attained the level of technological advancement required to create simulations of entire universes (similar to ours).
If (1) is true, then by definition neither (2) nor (3) can be true (since simulations would effectively not be possible). If (2) is true, then (1) is not true by default, and (3) would not be true as it would not be likely that we are in a simulation since it is not likely that a simulation exists. If (3) is true, then it is clear that both (1) and (2) are false.
Now, what about the premise that they are all false? Well, that is incredibly unlikely, and this is why. Suppose that both (1) and (2) are false. Then, there exists a sufficiently advanced civilization which can create simulations, and that civilization is interested in doing so. If that is the case, then that civilization would have, in all likelihood, created such a simulation in which billions of beings would be simulated. In fact, it is likely that many thousands of simulations are being run. If this civilization is sufficiently advanced (and we assume it is), then there would be a much greater number of simulated universes than "real" universes (of which there would only be one). By extension, there would be a much larger number of simulated "minds" than real minds. Therefore, it is a very real probability that our minds are simulated.
What about the prospect that all 3 are false? Well, Dr. Bostrom argues that it is incoherent to think in such a way. I, personally, would argue against that; indeed, what if we are the most advanced civilization ever? What if we would father all other simulations? Only time can tell, but it is a very real possibility.
On this note I leave you, dear readers. Stay cool,
- Snowman
Indeed. Bostrom's third premise creates a very jarring leap in logic, to my eyes. It is possible that "Sufficiently advanced civilizations which can create simulations CAN exist, but do not yet".
ReplyDeleteBut taking one step back, his argument creates interesting speculation for any who think it is possible to have such a reality. As statistical analysis says we have no reason outside of faith to believe we are the first, it is statistically unlikely. I would suggest this is the reason premise three says "We are most likely" rather than simply "We are".
However, one further argument remains. Does the feeling that we have "free will" and the concept of being part of a simulation necessarily clash? I will explain that my position of the potential for AI changed, recently, by thinking this through and resolving the paradox. I think that conscious willingness to act in accordance with a set of rules would create a sense that there is no free will when, in actuality, there is one. Nonetheless, the reality we create would be called a simulation by the mere appearance of causality. It could be that all matters of causality, insofar as chaos theory and quantum physics suggest necessary randomness and emergent patterns, are just such an illusion.
In this case, a linguistic game is created. Simulations exist, yet each simulation is a reality. No reality could exist in which simulations cannot be created simply because all realities share their rules with what they create.
What productive purpose does this argument even serve? Well, if you are like me and feel that we have some transcendent knowledge of our own freedom, any convincing AI would shatter your world. Further, the advancements of psychiatry might seem to point in the direction against our knowledge and towards a simulated world. My suggestion is that these things do not violate our knowledge because this apparent paradox can be resolved by believing the rules that govern us are willingly followed.
*note: I don't go into the complexities her but, if you see the problem, I would answer "yes, this suggests matter has some form of consciousness and that our minds (and bodies) are a unity formed of a community in constant argument".