Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Uploading minds into a simulated universe

  1. Sep 20, 2007 #1
    Could we ever reach a point where the entire universe can be simulated by computers, a la The Matrix? If so, could people ever 'upload' their minds into these simulations by, for example, replacing neurons one-by-one with artificial silicon components, until they are basically living (forever) inside the simulation?

    Because this has been causing me a bit of existential distress recently.
     
  2. jcsd
  3. Sep 20, 2007 #2

    Chi Meson

    User Avatar
    Science Advisor
    Homework Helper

    It's still science fiction now, because there is no scientific model (as far as I know) for such a thing. We don't really know yet how the brain works as a whole. It's more than just the neurons; it's the dynamical interactions among the neurons that make up our identities. Such chaotic interactions are kind of like the weather.

    I'd say, that right after we are able to predict weather patterns (exactly) a year or more into the future, we would be able to "program" an identity into a machine. I imagine it would be a copy of an identity, but it wouldn't be the "you" that was programed into it, even though it would believe that it was the "you", what with all your memories and all, but the original you would still perish with your body.

    Who put the stench in existential?
     
  4. Sep 20, 2007 #3
    http://en.wikipedia.org/wiki/Simulated_reality
    All you can do, really, is enjoy the ride.
     
  5. Sep 21, 2007 #4

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Apparently this point is not so clear when we talk about exact copies, but then Heisenberg probably rules that one out anyway [unless you happen upon some Heisenberg Compensators at a Star Trek auction]. And even then, I would never get into a transporter.
     
    Last edited: Sep 21, 2007
  6. Sep 21, 2007 #5

    Danger

    User Avatar
    Gold Member

    Ha! I was wondering where this thread went. I posted the first response to it, and when I submitted it I got an 'invalid thread' message.

    No, for the same reason that there can't be an omnipotent being. In order to model the universe, the computer would have to be larger than the universe. Where would you put it?
    It would also have to contain a model of itself, which is pretty ludicrous.
     
  7. Sep 21, 2007 #6
    this is incorrect- first there are several physically possible in-principle methods for infinite hypercomputation such as computation in Malament-Hogarth spacetime or infinite state Turing machines including possible kinds of quantum computers-

    but even an infinite universe is compressible if it is generated by a finite algorithm- that is if it has consistent rules/ locality/ Time- like ours- information about our world is limited to local frames of reference and bounded by locality/entropy which gives us the holographic principle- and QM says that all information outside that local bound is in superposition- not causally connected so not defined or 'computed'

    but even IF a universe's states were acausal and infinite- it could be generated by finite computation in principle since the cheapest algorithm to run ANY universe simple or unbounded in complexity is an algorithm which computes ALL possible universes that can exist- then to extract the desired history as needed- and regardless of how much information is expressed by some state of some universe even a small computer can render all it's possible relations and thus all histories given enough TIME- and computer time is invariant of simulation time-

    "In general, computing all evolutions of all universes is much cheaper in terms of information requirements than computing just one particular, arbitrarily chosen evolution. Why? Because the Great Programmer's algorithm that systematically enumerates and runs all universes (with all imaginable types of physical laws, wave functions, noise etc.) is very short (although it takes time). On the other hand, computing just one particular universe's evolution (with, say, one particular instance of noise), without computing the others, tends to be very expensive, because almost all individual universes are incompressible, as has been shown above. More is less!"

    Jürgen Schmidhuber
    http://arxiv.org/abs/quant-ph/9904050
     
    Last edited: Sep 21, 2007
  8. Sep 21, 2007 #7

    Astronuc

    User Avatar

    Staff: Mentor

    Beats worrying about getting a date on Friday or Saturday night, doesn't it? :rolleyes:

    Read the Hitchhiker's Guide to the Galaxy. It's not about the answer.


    Danger is quite correct.
     
  9. Sep 21, 2007 #8

    EnumaElish

    User Avatar
    Science Advisor
    Homework Helper

    Yes, that point has been reached decades ago. This is pretty much what each and every computer program does -- simulate the universe under (very) simple rules and (very) locally. Video games are a relatively recent example. What is at issue is the extent of the reality represented by the simulation. [I daresay a simulation that can faithfully represent 1/10,000th of the Reality would appear God-like.]
    Downloading the simulation into the mind would be easier. There is no reason why a person cannot be equipped with a network card type of a port and communicate with the simulation server (and other "slaves" --pun intended) on the local network. I thought the construct of the Matrix is more similar to this than the other way around, but I may be wrong.

    I empathize. I used to get stressed because (well, other than it's Friday night) of the teleportation problem. (How do you teleport the sense of continuity?) (And, does the original have to be destroyed in the process?) (Don't destroy the original -- at least not before checking that the new model is alive.) (If possible, don't destroy the original under any circumstances.) (How's that not murder?) (I WON'T LET YOU DESTROY ME, YOU MURDERERS!) (Even if my "being" is built quantum by quantum at the other end, these are not my quanta. And I think that matters.)
     
    Last edited: Sep 21, 2007
  10. Sep 21, 2007 #9

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    :rofl: So then you have resolved this issue all by yourself?!?!?
     
  11. Sep 22, 2007 #10

    DaveC426913

    User Avatar
    Gold Member

    You know, that's actually a pretty good argument.

    If a being is omniscient, does that mean its storage capability would have to encompass the universe? And if so, where would you keep it?
     
  12. Sep 22, 2007 #11

    Danger

    User Avatar
    Gold Member

    The idea is not original to me, but I totally agree with it. I believe that it was one of the SF giants who brought it to my attention, but I can't remember which one. Might have been Heinlein, David Gerrold, Phillip Jose Farmer...? In any event, I was a little kid at the time. I'm pretty sure, though, that I would have come up with it on my own in time.

    SetAI, you obviously have a lot of computer savvy, but I consider your arguement specious. In order to model the universe, you have to include every single particle, as well as virtual particles and other vacuum fluctuations, as well as the vagaries of thought processes within the inhabitants of that universe. Don't try to tell me that even the best supercomputer on the planet can model itself, let alone anything outside of itself. Where would it possibly find the memory capacity or processor speed to track every single quark and lepton in its substance, let alone every single electron moving through its circuits and every photon given off as IR through fricitonal losses and every quantum tunnelling event that results in cross-talk among closely packed circuits? It ain't gonna happen.
     
  13. Sep 22, 2007 #12

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    That's fine, but I hardly think anyone here or anywhere is qualified to make definitive statements.
     
  14. Sep 22, 2007 #13
    Okay, this is kind of a reference to another thread I started some time ago about artificial brains. Could we, at any point in the future, come up with the technology to replace the brain's neurons one-by-one until you are actually living inside the simulation? I'm not talking about copying consciousness. I'm asking if we could ever come up with the technology to actually move one's consciousness into a virtual universe.

    Is there some fundamental reason why this is not possible?
     
    Last edited: Sep 22, 2007
  15. Sep 22, 2007 #14

    ZapperZ

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    What exactly is meant by "simulating" the universe? I mean, what do you use as a starting point? Do you start at the individual particle interaction (vanesch already knows where I'm going with this), or do you already start at collective interaction?

    If you start with the former, take note that as of NOW, there has been zero evidence and ability to derive emergent phenomena such as superconductivity from considering all the relevant interaction at the single-particle scale. This means that there is no guarantee that even if you know and simulate all the interactions at that level, you'll discover all the phenomena that involves higher order collective effects.

    Also note that even with the 3-body problem, we still do not have a complete solution for the most general situation. So forget about the universe or even superconductivity. If you can show that you can solve for the 3-body problem completely, then I'd say you can start arguing for modeling something more complicated then that. Till that can be shown, I'm skeptical about any such claims.

    Zz.
     
  16. Sep 22, 2007 #15

    DaveC426913

    User Avatar
    Gold Member

    For starters, we do not yet know what consciousness is. We presume that it is the collective interactions of billions of neurons, but who knows whether consciousness itself would be translated into another medium?

    Raises the question: if we did simulate a mind, how would we know it is conscious? There is no test; the Turing test only tests for intelligence.
     
  17. Sep 22, 2007 #16

    Danger

    User Avatar
    Gold Member

    Good points, Dave.
    I have my own thoughts regarding the matter, but they're based solely upon semi-educated guesses.
    To start with, we retain our sense of 'self' despite the fact that the cells of our bodies, and even the atoms themselves, are constantly being replaced by new ones. I doubt that there's one single atom left in my body that was a part of me when I was born.* In that light, I think that if the engrams of someone's brain could be successfully duplicated in a computer, that 'model' would indeed be conscious. That wouldn't present a problem if the original is destroyed. If, however, one or more duplicates are made, whether or not the original survives, each would think that it's the real 'person'. The restriction of movement, perception, whatever imposed by being part of a machine would be definite evidence of its true situation, but it would still have that sense of 'self' as well as all of the original's memories up to the time of duplication. From that moment on, they would all become distinctly different 'people' due to differing experiences. If the biological unit no longer exists, who can say which is the 'real' one?

    *I saw written somewhere that with every breath you take, you inhale several million (or billion?) atoms that were once part of Julius Caesar's body. The same can be said for anyone else that has lived on the planet long enough for those atoms to have dispersed completely. In that sense, we all still have a lot of our original atoms, but not likely in the same role that they played the first time around.
     
  18. Sep 22, 2007 #17

    DaveC426913

    User Avatar
    Gold Member

    I like to pursue similar ideas. Robert J. Sawyer wrote a book called The Terminal Experiment wherein the main character copies himself to a computer. His simulation is turned on and has a few moments of disorietentation, thinking something has gone wrong and the experiment failed; he cannot see or hear. He calms down once it is pointed out that he is, indeed, a copy inside a computer.

    But I think RJS missed an opportunity here. I think the simulation would not accept the fact that he's "just a simulation" so easily. I think that the identity in the computer would interpret itself as the original. I think he would realize in a flash the answer to the question "What makes me me?"
     
  19. Sep 22, 2007 #18

    EnumaElish

    User Avatar
    Science Advisor
    Homework Helper

    I was assuming that the OP would settle for a passable copy. Otherwise, I agree.

    Besides, if we don't know what comprises 95% of the universe, then simulating it in some "exact" sense sounds like a tall order.
    Call me a chicken, but I'll walk into the Transporter/Copier Apparatus apres vous.
    ...
    ...
    ...
    ...
    (Where's that chicken!?)
     
    Last edited: Sep 22, 2007
  20. Sep 22, 2007 #19
    No to much to say about the topic but i found this link that could be interesting:

    http://www.spacedaily.com/news/robot-03o.html
    Artificial Development To Build Biggest Spiking Neural Network
    Artificial Development, Inc. today announced that it has completed assembly of the first functional portion of a prototype of Ccortex, a 20-billion neuron emulation of the human cortex, which it will use to build a next-generation artificial intelligence system
     
  21. Sep 22, 2007 #20

    Danger

    User Avatar
    Gold Member

    Cool link. Thanks.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Uploading minds into a simulated universe
  1. Mind (Replies: 9)

  2. The mind (Replies: 4)

Loading...