Sister blog of Physicists of the Caribbean in which I babble about non-astronomy stuff, because everyone needs a hobby

Saturday 5 January 2019

Infinite knowledge wouldn't enable all predictions, apparently

Interesting, but I think I'm missing something. In a deterministic universe with all laws and initial conditions known and unlimited computing power, I can surely just compute which light bulb I'm going to turn on in advance of the 11:55 prediction moment. Thus I don't really have any kind of free choice at all, just foreknowledge of what I'm going to do. That the machine responds based on my apparent choice seems irrelevant.
https://aeon.co/videos/if-you-knew-everything-could-you-predict-anything-a-thought-experiment

10 comments:

  1. Yeah I don't follow this. It seems like it presupposes that the future can't be predicted. Despite actually knowing the future you're arbitrarily required to report your guess in a way that by definition has to be opposite of your prediction. That doesn't show anything other than that you shouldn't make your prediction through that method.

    I think the scenario this is trying to replicate is if you know the future, and you're talking to your friend, and your friend would absolutely try to do the opposite of anything that you predict they'll do. You tell them they're going to say "hi" in exactly 5 seconds. So what stops your friend from saying nothing or saying "you're a freaking idiot" instead? I'll admit it's a hard question, but your prediction of the future would account for the fact that your friend would hear the prediction. It doesn't seem to me that you can rule out some kind of equilibrium prediction that your friend would hear and be unable to avoid doing. Assuming your friend will absolutely say the opposite of an infinite number of possible predictions seems like begging the question.

    Or maybe this just means you write down your real prediction in secret, lie to your friend about your prediction, and then show them the real prediction after the fact. If anything all this light bulb scenario says is there may exist some methods of communicating your predictions that are impossible to use correctly.

    ReplyDelete
  2. Of course you get absurd results with absurd basic assumptions. We don't know everything. The universe is not deterministic just because someone thought it might be. There's only one deterministic possibility and infinite others, so the probability of determinism is one to infinity. The burden of proof is with those who claim it is deterministic, and they can't demonstrate it. End of story.

    ReplyDelete
  3. Very interesting. As you may know, Predestination and Free Will is a debate of long standing within Christianity. I was talking about it with my father, the ordained minister, who was reading his newspaper when I asked him about the issue.

    He peered out from behind his newspaper, looked over his bifocals and said "It's a nonsense question. If you're an eternal being, time is just another dimension. You'd know the beginning and the end of every story. For human beings, who live in time, to even worry about free will - is nonsense."

    My Dad didn't believe in Free Will as everyone else supposes it, even on a non-theological basis. We make decisions gradually, one piece at a time, one aspect fitting into another one until at last the decision seems to be made - we choose the college, we buy the house, we marry the girl. As that happy-crappy old saying goes, "It's not the one big decision, it's all the small decisions we make which matter."

    ReplyDelete
  4. Ryan Beck Yes, I think communication is exactly the issue. The status of the prediction bulb cannot be the same as the researcher's actual knowledge, regardless of whether they get a sneak peek at the computer's prediction. As soon as the prediction bulb is set, the researcher knows what the second bulb will do. Being forced to "say" the opposite via the prediction bulb doesn't mean they really believe that's what's going to happen. The machine is clearly a "frustrator" but only in an everyday sense of being annoying, not because it makes foreknowledge impossible.

    I would imagine that in the scenario of telling people what they're going to do in a purely deterministic scenario, you don't have any choice as to what you're going to say either. So you'd get a staple of sci-fi show conversations which goes something like this :
    "I can predict the future. Ask me how."
    "OK, how ?"
    "I knew you were going to say that..."
    And so on. In essence, you'll never ask them to do something you know they won't do. You can't do that, by definition.

    ReplyDelete
  5. Sakari Maaranen I think this was supposed to be an attempt to say that we can't know everything even in the hypothetical case of a deterministic universe. A better argument for that would simply be that this would require infinite precision, which is impossible.

    ReplyDelete
  6. Rhys Taylor a related point is that the simplest possible deterministic universe is the one with nothing in it, aka. nothingness. It can't exist even if it did. Any other starting point you may choose would be an inexplicable, arbitrary complication that would then determine the entire universe. It seems like a big jump from nothing to everything. Determinism does not allow other scenarios. The initial choice would not be deterministic, so a deterministic universe can only be the nonexistent one.

    ReplyDelete
  7. Rhys Taylor I have this theory about why Determinism still lingers on: LaPlace's demon is lying. It couldn't possibly know the location of everything . Ask the demon for the location of every electron and he'll run away, terribly embarrassed.

    ReplyDelete
  8. I think there has to be a difference between deterministic and predictability. A deterministic universe run according to unbreakable rules could still be unpredictable due to imprecision, complexity, and the (under-rated) difficulty for any inhabitants to determine what those rules are. Only magical demons could be privy to the infinite knowledge necessary to predict events infinitely far in the future, though the world would still be deterministic in the classical sense.

    A predictable universe need not be deterministic either, e.g. it might be possible to only make statistical, probabilistic predictions rather than exact ones. The underlying rules could still be unbreakable despite the element of randomness.

    I'm not saying the real world is like either of those scenarios, however.

    ReplyDelete
  9. Rhys Taylor Well, that's an excellent, king-sized caveat you've stapled onto the end of that comment. Seriously. Attempting to sum up the Real World in a deterministic mode - problem is, it's true, in a limited sense - and at certain scales we can depend on it. But at the edges of it all, those little butterflies in the Amazon will continue to flap their wings....

    ReplyDelete
  10. A computer and it's memory is a completely deterministic universe; and we know via the halting problem that there isn't a program that would be able to read the memory in it and predict the behavior, in general, of the system.

    ReplyDelete

Due to a small but consistent influx of spam, comments will now be checked before publishing. Only egregious spam/illegal/racist crap will be disapproved, everything else will be published.

Whose cloud is it anyway ?

I really don't understand the most militant climate activists who are also opposed to geoengineering . Or rather, I think I understand t...