Timothy asks:
“Are we in a simulation?”
Though it is not strictly a logical approach to ask a philosophical question by reflecting on a feeling I do feel it helps to add meaning and perspective for the question. I wanted to ask what are the chances that we are living in a simulation. One reason I ask this is because of the infinitely unlikely possibility when looked at scientifically that I would ever be alive and yet less to be a human and in the era I could write this. Either this is like winning the lottery 1 million times in a row or something else is at work that I am alive, as a human (my opinion the best thing to be), in this era with all this technology. Thoughts?
Answer by Craig Skinner
Wonder at your existence, worry that you might be a simulant. Two philosophically interesting thoughts in one question. Also, it’s just fine if a question arises from a feeling.
Yes, your (or my, or my cat’s, or the squirrel in my garden’s) existence was hugely improbable. Had a different sperm out of the millions competing to penetrate the ovum been successful, had your father been away on business on the day you were conceived, you wouldnt exist. Also the huge fluke that your parents chose to mate with each other rather than with one of the many alternatives. And it’s mind boggling to think that not a single one of your millions of forebears over 3 billion years failed to reproduce. If just one of your myriad fishy ancestors had been eaten by a bigger fish when young, no you. And yes it is like a lottery. Just as somebody has to win however tiny the odds of success, so, given that you exist, you must be somebody, and somebody has to be you.
As to whether you (and I) are simulants, the answer is it’s a distinct possibility but it’s difficult to assign a probability. We assume there is a real world with intelligent beings in it. And your question is whether you could be a simulated being in a simulated world created by these real beings who run the simulation on their supercomputers.
Clearly such technology is beyond current human capabilities. But given the pace of developments in computing and AI, it may be that future superintelligences, either augmented humans, or maybe more likely nonhuman following the singularity of the development of smarter-than-us AIs which rapidly engineer ever-smarter successors to reach unimaginable levels of intelligence.
And so the probability you seek depends on:
- Whether superintelligences will arise.
- Whether they would be interested in simulating earlier eras of human life (say as a game or for historical research), rather than ignoring or eliminating humans.
- Whether such simulations would be frequent (like our computer games played in most households).
- Whether simulated humans would be genuinely conscious (like you and me).
If the answers to these four questions are all yes, and they might be, then right now it’s far more likely that we exist in a future simulation than that we live in the real world in the 21st century.
So, in case we do, let’s tell the Simulater that we’re on to it:
Hey there BIGBRAIN, we know you’re there. And in case you had doubts, yes we are aware, we love, laugh, cry and care about our little lives down here. And dont get too smug up there in your real world, for you and your world could themselves just be simulations by yet higher orders of intelligence in the proper real world. Or maybe it’s simulations, rather than turtles, all the way.
Ask a Philosopher 
Thanks for pointer to your stimulating and informative 3-part piece. It does seem as if it would be cheaper just to duplicate the real thing than to make a sufficiently sophisticated simulation. But before concluding I am not a simulant, can I ask 2 questions. First could quantum computing, embryonic but now showing some promise, move the goalposts. Secondly, is it necessary to program in every last detail of all (simulated) events. Would not setting up the initial state comprising fundamentals, their powers and their aggregation into things and creatures, plus inputting rules for transition (laws of nature) allow the simulation to run. No need to program a separate instruction to each photon to move at light speed, they would all just do so; no need to input the very detailed info to allow an individual protein to fold in its unique way, they would just do it according to their nature as in reality. I suppose what I’m saying is that the simulation would be a giant cellular automaton, but then maybe the real world is too.
First about the question whether quantum computing can move the goalpost, as you call it. I am not an expert on this field, but I dont think so. I am citing https://en.wikipedia.org/wiki/Quantum_computing (the section on “Relation to computational complexity theory”: “The class of problems that can be efficiently solved by quantum computers is called BQP, for “bounded error, quantum, polynomial time” […] The class of problems that can be efficiently solved by quantum computers is called BQP, for “bounded error, quantum, polynomial time”.” The article contains a graphic representation of this (the suspected relationship of BQP to other problem spaces: https://en.wikipedia.org/wiki/Quantum_computing#/media/File:BQP_complexity_class_diagram.svg.
So probably, quantum computing would not do the trick. Moreover, there is the coherence problem. Quantum computers must be shielded from outside influence. I don’t see this becoming practically feasable for problems involving large amounts of data. I think the technology will have some niche applications, but that’s it.
A cellular automaton is a computer with a high degree of parallelism. However, you cannot make it large enough to calculate NP-hard problems except for simple cases with small numbers.
Secondly, I suspect that phisics is not Turing computable, i.e. every formal theory of reality will turn out to be incomplete or at least computationally incomplete, see https://creativisticphilosophy.wordpress.com/2017/11/07/computability-and-physics/
If I am right, an accurate simulation of reality is not going to be possible, no matter if you are a super civilization or not. That is a mathematical impossibility, not a technological one. So in order to answer the question if we are living inside a simulation we have to look for phenomena that are not Turing computable, i.e. every formal theory about them would be incomplete.
If we are living inside a simulation and it is just an approximation, we should be able to spot inconsistencies somewhere.
I think a simulated human being would indeed be conscious, I have explained my argument towards this opinion here: https://wordpress.com/post/creativisticphilosophy.wordpress.com/619.
However, I think the formation of a super civilization is extremely unlikely. I think technical civilizations like our own are typically short lived. You are going to find my ideas about this here: https://embassyofthefuture.wordpress.com/2015/04/18/civilizations/ and here https://embassyofthefuture.wordpress.com/2015/07/12/being-strangeled-by-the-invisible-hand/. Moreover, I have reason to believe that superintelligences are impossible, see https://creativisticphilosophy.wordpress.com/2016/06/25/is-the-singularity-near/. On the same website, you are going to find some articles containing some arguments for the opinion that human type intelligence can be done by physical systems but not by algorithms (so a simulated mind might be conscious but it will have a limited intelligence). In this view, truely intelligent beings cannot be described completely by any single formal theory or algorithm. They are “proteons”, not “systems” (this is also the deeper reason for the science/humanities divide and for the persistence of philosophy.
I hope this comment will be allowed here although it contains so many links, but I believe these links are relevant for the question.
Kind regards
Andreas Keller
An important aspect here is the question wheter such simulations are possible. I have serious doubts about that and I have expressed those doubts here: https://philosophicalexcavations.wordpress.com/2017/07/30/are-we-living-inside-a-computer-simulation-part-1/ (and in the two following articles).
Kind regards
Andreas Keller