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

Experience Machine

  1. Sep 20, 2004 #1
    Now that I am fully immersed in my Philosophy 110 resitation class every friday, I tend to leave class asking many questions about the discussions we had during class. We were discussing something known as the 'experience machine' and the scenario is as follows.

    Hypothetically, there is a machine which you can connect yourself to which simulates virtual reality so real that you could not tell the difference. Once you are connected to this machine, you do not know that you are in virtual reality, and you assume it is your real life. This machine guarantees more happiness than the current life you have, and by hooking up, you do not change the 'real' world in any way from what it would have become if it did not experience your absence. So, would you decide to hook up?

    To me, It would be dumb not to. But many in my class, including the TA, said they would like to remain in the real world because the successes you would have in the 'real' world would somehow be of more value.

    My logic is that if you are not helping or hurting the world you are leaving, and you are guaranteed more happiness, than hooking up would be the most logical thing to do; that is, if you believe that happiness is the ultimate desired end.

    The feeling of some of my classmates was that your accomplishments in the virtual world would not be as valid as the ones you could make in the real world, but I couldnt understand why the validity of one's accomplishments would matter as long as one didnt know they were less valid in a virtual world (remember, once in the virtual world, you do not know of the real world and assume the virtual world IS the real world, and thus dont know that your accomplishments arent valid).

    I was really struggling in class trying to understand the logic of their side, and my TA made an analogy of a red barn, one that doesnt exist but that I can see. He said that the class fealt that I should be told that the red barn doesnt exist by an objective source who could see the truth. But I really dont think there is any logic in notifying me of its existence or non-existence as long as my perceptions arent going to hurt anything else, just like the virtual world.

  2. jcsd
  3. Sep 20, 2004 #2
    This is a very used up topic i think. But here are my thoughts on it.

    I can understand why your classmates would say that your accomplishment are not as valid if in the Virtual world as if in the Real World. The only way they are less valid is that they will not be remembered by anyone, and no one will actually no of them.
    Personally i would not hook up, im here for life, wether its the most painful thing or bliss, im here for life. Not to live in a virtual life. If you were to not know that you were in this world. Your only lying to yourself. Also you remove yourself from the real world to, therefore impacting other peoples real life. So i personally would stay.
  4. Sep 20, 2004 #3
    Reality check, you are already an experience machine. No machine in the universe now or ever will ever be able to do that(give you more happieness). I speak it now an any human who ever goes against these words will be wrong. You will never be able to have more happieness than the sum of your present system. You might be able to throw the switch for longer periods at the expense of your future. They sometimes call that "drugs"(many kinds of drugs). There are also other ways, but you would destroy the human being in that way. Virutal world is a destructive one. It does not know what you need, but will condition you to a way which is most likely not what you need. You will not know it was happening, but it would. It even happens now for you.
  5. Sep 20, 2004 #4
    TENYEARS, i believe your wrong, im sure in a few hundred years scientist will know enough about the brain to stimulate you being in ultimate bliss continuously...
  6. Sep 20, 2004 #5
    you can do that now with drugs. why wait?

    ah, you intuitively know that you would be wasting your body. Why have a body if you are not going to use it to 'fully experience' this reality??

    olde drunk
  7. Sep 20, 2004 #6
    I feel that taking away my POTENTIAL for increasing the greatness of the world would be the leading reason as to not hook up to this machine. I believe the fact in knowing that the happiness in my life is either earned, or randomly distributed (like finding 20 bucks); I would feel more "satisfied" with my life. More content. I don't think that would be possible knowing I was in a machine.
  8. Sep 20, 2004 #7
    It matters not what you believe, with the expecption that what you believe sometimes guides the path of your life. Careful what you believe. I do not believe what I have spoken. I know it. This type of belief is what will ultimately destroy most of humanity. Believe what you wish young and old.
  9. Sep 21, 2004 #8
    That sounds good, but really, why would it matter if someone was remembered or known of? It doesnt matter when you are dead? Right? And as stated before, the validity would NEVER be known by the person in the machine, and so, he would not be aware that he will not be remembered.

    This is a strictly hypothetical scenario, Im not saying it is a concievable idea.

    You wouldnt know you were in the machine, it was one of my original premises.

    So you value other people's quality of live over your happiness?
  10. Sep 21, 2004 #9
    Questions, oh dear lord questions...

    This 'Experience Machine'... Would it be made to have uncertainty? Would everything be certain? (programmed). You would control your 'virtual life' with your real brain, wouldn't you? So your choices would not always be certain there either? I believe that it would be up to the person who is honestly asked this question to experience this for real to decide whether or not they would do it.

    As long as you could interact with other real people, and make it in an entirely different civilization of good/bad deeds, wants, desires, life style, etc I would do it. Would this experience machine 'control' your brain? Else how wouldn't you know you were in a machine? How would you use your 5 senses? Reproduce? Would all of these sensations (yes, I'm also talking about the feelings experienced during sex) also be felt within this virtual world?

    What I would miss is the vastness of real. If you could get all 6 billion people on the Earth to use this virtual life as their real life than all of what you did in this virtual world would be taken the same way as if you did it in the real world (in responses, interaction, butterfly effect).

    Imagine how important a teacher could potentially be to a child. But this teacher is in a virtual life, where this child is not.

    How would this world truely be any happier than this one if you have real minds controlling the environment?

  11. Sep 21, 2004 #10
    What do you think the world is comming to? It will never be able to do what you say in terms of more, but the world is attempting to control you in ways of when by creating pathways of conditioned response. The world has gone mad and does not quite know it yet.
  12. Sep 21, 2004 #11
    i understand where your coming from man id rather be any where else thats why i space out randomly lol
    but my dreams are fantasy of course i hate the real world but i wonder if things would be diferent if my life had'nt gone the way it did dont ask im not telling​
    Last edited: Sep 21, 2004
  13. Sep 21, 2004 #12
    From my experience, my life was the result of my actions and choices. Once I accepted that responsibility, I had the power to change my life to the direction I wanted.

    Several philosophers have said that when you point a finger it always ends up pointing at yourself.

    No matter what, you have all your answers and can change your life. Believe in yourself and you will be amazed at what you can do. No BS!

    olde drunk
  14. Sep 21, 2004 #13
    Wait for it.... Wait for it..... What if we are already in such a machine and all of what we think of as reality is but illusion!!!! (someone had to!) :)
  15. Sep 22, 2004 #14
    the very essence of hindu philosophy, if anyone here knows, is that the whole world is an illusion. Attainment of "Moksha" is considered as eternal bliss and as a disentanglement from the birth and death cycle.
    so as someone quoted....we are an experience machine in ourselves, living in an "illusionary" world. each individual perceives the world in his own way. if you want permanent happiness, there are two ways:
    1) perceive every failure as a success in this illusionary world and celebrate.
    2) strive for attainment of "moksha" so that you will be eternally happy after your incarnation here on earth.
  16. Sep 22, 2004 #15
    Of course, to an extent. Aren't most people willing to make sacrifices to help others?
  17. Sep 22, 2004 #16
    I have a problem with your professor offering that hypothetical at all. He or she obviously intends to discuss some kind of value theory or moral theory concerning it. However, such a machine is not possible; no machine will ever simulate reality exactly because that would require computation outside physical possibility.

    I believe that trying to determine morals based on a situation that is impossible leads to faulty conclusions. It would seem to me that since morality is entirely based on the world around us, deriving value or moral judgements from a situation that is not based upon the world around us would lead to invalid conclusions.
  18. Sep 22, 2004 #17


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    A virtual reality machine that produces a system of sensory inputs indistinguishable from the kind we encounter in the actual world isn't a physical impossibility. Such a machine would not need to simulate the world down to the level of atoms. All it would need to do is simulate a world in sufficient detail so as to meet the human sensory limits. Any level of detail beyond these limits wouldn't add anything to the apparent 'real-ness' of the virtual environment, and so could be omitted accordingly.
  19. Sep 24, 2004 #18
    I dissagree completely. To provide sensory input to a human is something I think would be entirely possible. To make that sensory input replicate our world to such a degree that it was indistinguishable from the real one is completely impossible. You say you wouldn't have to simulate the world of atoms; but if you didn't this world would be inherently different from the real one.

    If you do not simulate atoms, then when I break something, such as glass or metal, it will not break in the same way as the real world. Since it will not react the same, huge differences with the real world become apparent; how you build a bridge, how a building collapses, etc.

    Since you cannot simulate the real world exactly, you can never provide a world that is the same, and therefore you can always think of a reason one would rather be in the real world. Trying to explore moral values of an unreal system is interesting, but not useful or productive.
  20. Sep 24, 2004 #19


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Newtonian physics did a pretty good job of describing how the world behaves without mentioning atoms, don't you think?

    Even if we grant that such a simluated world could not behave perfectly like the real world, our case is not yet closed. All it needs to do is produce input that is consistent with a human observor's expectations of how the world will act, and the fact of the matter is that humans do not have such a powerful intuitive grasp of physics that slight deviations from what should have happened will be conspicuous.

    For example, there are many ways that a glass could be shattered without raising a human's suspicions. When a glass shatters, do you meticulously notice the exact position and shape of each shard? No; you notice a general pattern of how the shards spread and how they tend to be shaped. If one of the shards is not shaped precisely as it should have been, you will not suddenly become skeptical of the reality of your world. It's questionable that even our most detailed physical modeling of a glass-shattering system could capture exactly the outcome, because of all the small variables involved. And it's pretty clear that a human's intuitive grasp of physics is not as exacting as our most precise mathematical models in physics are.

    Even if we grant that an event in the simulated world is so grossly out of proportion with what should have happened that it raises the suspicions of a human, will this cause the human to become a global skeptic who has concluded that he is living in a simulated world? Strange phenomena that don't seem to fit into our physical model of reality happen all the time in real life. This does not cause people to become skeptical as to the ultimate nature of reality. At most, what it causes them to do is try to find some way to rationalize how it could have happened within their model of how the world works. More likely, it will cause them to shrug their shoulders and move on with their life.
  21. Sep 24, 2004 #20
    I hope you don't think my dwelling on this point to be improper or a sign of being stubborn. I will admit, the more I think about this, the more I feel I'm correct (a sure sign of trouble to come?) but am also greatly interested in the conversation. You have brought up some great points.

    Absolutely not! Or, at least not without many advances and improvements. I would say there are some very basic things about the world around us that cannot be predicted with NMech even in the most rudimentary way without other major steps - for instance, the idea of the atom.

    True, I do not. However, I currently work in the laboratory at a materials engineering firm, and over time I've become accustomed to looking very closely at many kinds of fractures. Although I am not nearly as skilled as my boss, I've found he can tell a huge amount of information after studying a fracture for a short amount of time. Anything that breaks gets studied. These fractures occur at a molecular level, so any fake-world would need to contain algorithms for molecular interactions to fool him.

    This leads to several questions. Who would you design this world to fool? My boss, me, or someone else? If it be someone else, who might not notice a "fudge factor" in regards to crystal lattice, what might they notice that I would not? Must you make it to fool everyone, and how would you do that without simulating everything exactly?

    Of course, this is hypothetical, so I suppose you could say that you are just creating this machine for one person. However, could you actually grasp everything? Your next point is a very good one I think:

    It might not make them a global skeptic, but it might change the quality of the life they are existing, which is paramount to saying this machine is not the same as being alive (an important part of the original moral quandry). For instance, my shins are faintly sore from Karate last tuesday. Not a painful sore, just enough to feel if I knock on them. It doesen't hurt, but the sensation gives me a truly delightful adrenaline rush and reminds me I'm going back again tomorrow.

    Can the machine produce sensations so subtle? Can it possibly produce all of them? What about the color of a fire or hot sand between my toes? As the creator of the machine, you are faced with doing one of two things. Either one - as you suggest- you could use algorithms that produce about what people observe, without worrying about exactness. The problem with this is that you need a vast number of algorithms. Or secondly, you could try to build from the "ground up" by simulating quantum interactions and letting everything build from there. In the first case you end up with too many algorithms, and in the second too many calculations. This goes back to Feynman's findings on computation and the universe.

    I believe this moral situation given in the original post is based upon an important logical falacy. Any machine will be made out of things of this universe. How can something made out of the stuff of this universe simulate this universe in a reasonable way? If the machine cannot, in fact, be produced, what is there to be learned from the question?

    The two arguments I'm making are both of the more difficult kind of argument to make. First, that something can't be done, and secondly, that a question won't provide a useful answer. Have I been convincing at all, or should I save my fingers from more pounding? :tongue2:
  22. Sep 26, 2004 #21


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    It seems like there are a number of ways a simulated world could handle these contingencies.

    The most straightforward method would be to only simulate the world on the molecular level when needed. For example, the simulator may treat the world roughly as Newtonian most of the time, but locally switch to the quantum level when needed-- eg, when there is a scientific experiment whose results depend significantly on the distinction between the two models.

    Another approach would be for the simulator to simulate molecular level interactions statistically, rather than directly. For instance, rather than actually do computations for the interaction of every molecule in a cracked substrate, the simulator could do something like this:

    1. After experiencing conditions of type C, materials of type X tend to do develop cracks that follow so-and-so kind of distributions, with such-and-such means and standard deviations.
    2. This material is of type X.
    3. This material has experienced conditions of type C.
    4. Therefore, it should develop cracks that follow a so-and-so distribution, with such-and-such mean and standard deviation.

    An even more elegant approach would be to simply produce results that are consistent with the expectations of the observor. It may be possible that a machine could read one's neural code to the extent that it could deduce what kind of sensory input is expected by the observor under certain conditions. If it could do this, then your boss would never find anything at odds with what he expects to see.

    These solutions are all a little ad hoc and problematic, of course, since I'm trying to be at least somewhat economical. But since we're talking about possibility in principle, we could just as well go to the extreme and suppose that we build a simulator that simulates the entire planet earth on a quantum level. That's a pretty exorbitant simulation, but for the purposes of this thread we only need to suppose that it runs (in simulation-time) for the lifespan of a human participant-- about 100 years at most. Such a simulator would need vast computational and temporal resources, but probably not so much that they would outrun the actual computational and temporal resources available in the universe. Therefore, we can't rule out the possibility that a sufficiently advanced technology could actually pull it off.
  23. Sep 27, 2004 #22
    Now that is an elegant solution. The machine could use the person's brain to generate the sensory inputs. I see some serious problems with that, but you did too, as you mentioned you were being economical. I will let the issue go, for now.

    Thank you for your responses.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook