Newcomb's Problem

[Sommerswerd]

Seems like a subtly religious question...

 

[Sommerswerd]

On second thought... its not subtle... This question is basically asking whether you are "believer" or not.

Sorry if this angle has already been discussed. I didn't read the thread.

 

[TheMeInTeam]

On second thought... its not subtle... This question is basically asking whether you are "believer" or not.

No, it isn't. It's only "religious" if you look at it with bias in that lens. Humanity has many times encountered systems that worked in ways we didn't understand yet, or generated models that make better-than-random predictions despite not understanding everything going on behind the scenes.

In contrast with religious beliefs, "omega" is providing evidence. That evidence is crushing and consistent. We're not told how omega is doing it, and for the purposes of making a decision it doesn't matter if we hold that it behaves consistently on our choice. The model that most consistently generates more stuff (and we're given tangible evidence) is one-boxing, because omega is using SOME method of accurately predicting future choices.

There is nothing, in principle, that prevents this being possible without any special religious attachment.

It's not a probability guys, it's a time travel paradox, ie a casual loop.

You don't have any evidence that merits concluding that, given the scenario as described. "Omega" could be something like a super-accurate computer that can make observations and model human future choice accurately, or something else we're not picturing. It can then bat 100% or near 100% (near enough that it could reasonably be expected to get 100 straight predictions right) without so much as challenging our notion of causality.

 

[Sommerswerd]

No, it isn't. It's only "religious" if you look at it with bias in that lens.

Well, I look at it with "bias in that lens" (whatever that means)... so by your definition it is "religious".

Humanity has many times encountered systems that worked in ways we didn't understand yet, or generated models that make better-than-random predictions despite not understanding everything going on behind the scenes

or came up with religions to do the same thing... so again... religion is clearly implicated by the question. Did you pick "I believe in the almighty Omega" and now you have some buyer's remorse? Is that why you're bristling at this?... I'm gonna check what you picked. (I was going to pick "believe in Omega" until I realized the religious implications).

Anyway, just poking a bit of fun at you as this is obviously a lighthearded topic.

(EDIT: I checked your pick...called it!)

 

[Omega124]

Well, what criteria determines that they're not equally probable? The only factor that changes this sort of thing is Omega's omniscience. But the human doesn't know what he will choose until he's faced with the choice. How would you quantify Omega's omniscience?

Hi. I'm omniscient. :3

 

[TheMeInTeam]

Well, I look at it with "bias in that lens" (whatever that means)... so by your definition it is "religious".

Only insofar as typing at your computer is "religious". The assertion that one vs two boxing has anything to do with religion is tangential/absurd. There's no reason to even consider that.

or came up with religions to do the same thing... so again... religion is clearly implicated by the question

There is no analogy in the case you're making. Omega is observed evidence, not a belief in absence of evidence. People had religious beliefs and fabricated stories before we knew what cells and atoms were, but those beliefs/stories were not supported with evidence.

In this case we're presented a 100x in a row experimental trial and measurable outcomes. Show me a religion with those ^_^.

Did you pick "I believe in the almighty Omega" and now you have some buyer's remorse?

Is that a serious question? If we encountered this scenario for real I'd be up ~1 million dollars on you if you didn't also one box. That's not the position that feels remorse.

The position that "one boxing" is "faith" is backwards. Experimental evidence *soundly* demonstrates that you get more from one boxing. Two-boxing suggests belief in something despite seeing repeatedly that it leads to the inferior result...in other words two-boxers are so sure of their understanding of causality and of omega that they're willing to ignore overwhelming evidence. THAT's the position that's supposed to be rational?

It's silly to paint the obviously experimentally weaker outcome as the better choice. If I'm in this scenario for real, I'd try to understand how omega works, but I'd still one-box if forced to choose without knowing.

I would make that choice for the same reason I'd avoid smoking. I don't know the exact, body-process level reasons smoking is likely to significantly shorten my lifespan and quality of life. I'm not a physician. Despite not knowing that I've seen enough that the model "doing this is bad for you" holds pretty well.

Telling me that one-boxing = faith is like telling me that avoiding smoking = faith. That's frustrating because it's nonsense .

 

[Sommerswerd]

Only insofar as typing at your computer is "religious". The assertion that one vs two boxing has anything to do with religion is tangential/absurd. There's no reason to even consider that. There is no analogy in the case you're making. Omega is observed evidence, not a belief in absence of evidence. People had religious beliefs and fabricated stories before we knew what cells and atoms were, but those beliefs/stories were not supported with evidence. In this case we're presented a 100x in a row experimental trial and measurable outcomes. Show me a religion with those ^_^. Is that a serious question? If we encountered this scenario for real I'd be up ~1 million dollars on you if you didn't also one box. That's not the position that feels remorse. The position that "one boxing" is "faith" is backwards. Experimental evidence *soundly* demonstrates that you get more from one boxing. Two-boxing suggests belief in something despite seeing repeatedly that it leads to the inferior result...in other words two-boxers are so sure of their understanding of causality and of omega that they're willing to ignore overwhelming evidence. THAT's the position that's supposed to be rational? It's silly to paint the obviously experimentally weaker outcome as the better choice. If I'm in this scenario for real, I'd try to understand how omega works, but I'd still one-box if forced to choose without knowing. I would make that choice for the same reason I'd avoid smoking. I don't know the exact, body-process level reasons smoking is likely to significantly shorten my lifespan and quality of life. I'm not a physician. Despite not knowing that I've seen enough that the model "doing this is bad for you" holds pretty well. Telling me that one-boxing = faith is like telling me that avoiding smoking = faith. That's frustrating because it's nonsense .

I'm gonna put all your assertions of "silly", "absurd" and "nonsense" etc., aside, since they're not much more than subjective adjectives that don't tell us anything about the merits of the claim. What I will ask you directly is do you deny that Newcomb's problem/paradox, whatever, has legitimately arguably religious/theological implications? I mean a simple Google search shows that a lot of scholarly work has been done by authors, philosophers etc discussing exactly this at length. So I dunno TMIT, for you to unilaterally declare the test/question has no religious implications whatsoever seems... flat out factually inaccurate.

I mean I understand the position that it can be looked at from a purely mathematical/statistical/probability standpoint... but you seem to be saying it must be, which I disagree with. And TBH, even then, that requires some pretty big assumptions. For one thing... "Omega has been right 100 out of 100 times!" Says who? Omega? Or some guy purportedly speaking on Omega's behalf? And regardless of who said it do you believe it? Why? Did you observe all 100 people getting $1 million with your own eyes, or is this just "something that has been reported." In other words... are you taking it on faith? Why? Because its what you want to hear? Because you've already emotionally committed to the decision?

Or is it because the parameters of the test, indeed the universe we occupy as participants in the test requires us to accept Omega's prior infallibility as absolute truth beyond question? Any of this sound familiar?

 

[Manfred Belheim]

I'm gonna put all your assertions of "silly", "absurd" and "nonsense" etc., aside, since they're not much more than subjective adjectives that don't tell us anything about the merits of the claim. What I will ask you directly is do you deny that Newcomb's problem/paradox, whatever, has legitimately arguably religious/theological implications? I mean a simple Google search shows that a lot of scholarly work has been done by authors, philosophers etc discussing exactly this at length. So I dunno TMIT, for you to unilaterally declare the test/question has no religious implications whatsoever seems... flat out factually inaccurate.

I mean I understand the position that it can be looked at from a purely mathematical/statistical/probability standpoint... but you seem to be saying it must be, which I disagree with. And TBH, even then, that requires some pretty big assumptions. For one thing... "Omega has been right 100 out of 100 times!" Says who? Omega? Or some guy purportedly speaking on Omega's behalf? And regardless of who said it do you believe it? Why? Did you observe all 100 people getting $1 million with your own eyes, or is this just "something that has been reported." In other words... are you taking it on faith? Why? Because its what you want to hear? Because you've already emotionally committed to the decision?

Or is it because the parameters of the test, indeed the universe we occupy as participants in the test requires us to accept Omega's prior infallibility as absolute truth beyond question? Any of this sound familiar?

Lots of people have written lots of scholarly work discussing reality in religious terms. That doesn't mean any atheist has to consider any of those works valid or to have any actual bearing on reality. Likewise you can talk about Newcomb's problem however you like, but there's nothing inherently religious going on in the problem itself and so it's not a requirement to include it in any analysis you make.

As for who says Omega is infallible and why you should believe it - it's because that's the scenario presented in the question. The question isn't about how credible you would find that claim in reality, it's instead presented as a given for the scenario. If someone asks you what you would do if something is true then you don't need to have any faith that it actually is true in order to answer the question as if it were.

 

[Sommerswerd]

Lots of people have written lots of scholarly work discussing reality in religious terms. That doesn't mean any atheist has to consider any of those works valid or to have any actual bearing on reality. Likewise you can talk about Newcomb's problem however you like, but there's nothing inherently religious going on in the problem itself and so it's not a requirement to include it in any analysis you make.

I don't disagree with this, however you're reframing my point a bit, and then disagreeing(?) with the argument you made rather than the one I did. I am not asserting that it can only be looked at in religious terms. I am responding to TMIT's assertion that it cannot be discussed/imagined/framed etc in religious terms and to do so is erroneous and various other adjectives that basically mean "wrong".

As for who says Omega is infallible and why you should believe it - it's because that's the scenario presented in the question.

Which is exactly my point about assumptions. You are assuming that Omega has indeed been infallible for the purposes of the question. There are many other assumptions, you have to make, that is just one. The clergy tells you god is infallible and has always been so. If you are a believer, you assume that this is true for the purposes of the question "does god (Omega) somehow know ahead of time what is in your heart/mind and what actions you will take"?

If you assume the answer is yes, then you can believe that your action of picking the second box (follow the religion/accept Jesus as your lord and blah, blah blah, etc), thus proving your faith in god (Omega)'s infallibility and justifying your expectation that god (Omega) will have rewarded you by predetermining that this is exactly what you would do. Picking the second box is taking a leap of faith as to Omega's prior and continued infallibility.

 

[Manfred Belheim]

I don't disagree with this, however you're reframing my point a bit, and then disagreeing(?) with the argument you made rather than the one I did. I am not asserting that it can only be looked at in religious terms. I am responding to TMIT's assertion that it cannot be discussed/imagined/framed etc in religious terms and to do so is erroneous and various other adjectives that basically mean "wrong".

Fair enough, I didn't read the whole exchange in depth so probably got the wrong handle on it.

Which is exactly my point about assumptions. You are assuming that Omega has indeed been infallible for the purposes of the question.

Isn't that stated in the question? Just going from what Mr Wiki says:

A person is playing a game operated by the Predictor, an entity presented as somehow being exceptionally skilled at predicting people's actions.

If this were real life and something were presented to you in that way you'd be right to question it, but given that this isn't real life and is in fact a logic puzzle it doesn't really make sense to do so. The whole point of these sorts of logical thought experiments is that you take the premises as read and then analyse what logically follows from them. If you start questioning the premises themselves then you're going off on a tangent which, while possibly entertaining or illuminating in its own right, is not actually tackling the question. It's not a leap of faith to just accept the premises for the purpose of answering the (entirely hypothetical) question.

So sure you can turn it into an excuse to talk about trust, faith, credulity, greed, etc etc etc if you want to. But I think it's quite fair to say the the question is demonstrably not about those things and they are not relevant to trying to answer it.

 

[Omega124]

"Omega has been right 100 out of 100 times!" Says who? Omega? Or some guy purportedly speaking on Omega's behalf?

Quite frankly, I'm a bit insulted that you don't take my affidavit seriously. When have I ever lied?

OK, funposting aside, I voted for two boxes when I first found the thread, not really reading the whole OP. I've seen variants of this scenerio, with the only difference that the whole Omega bit was completely omitted from it. So you had two boxes, and you had to choose either a guaranteed 1000 or the chance to gamble for a million. And in that case, I still stick to the guaranteed 1000.

Having a being (which I'm going to call Alpha because I'm Omega ) with a proven, veritable track record does change the scenerio around. Honestly, I feel like having pre-knowledge of Alpha and their track record really mixes everything up, and truely there really is no winning scenerio for me.

So, people who picked Box B got a million dollars, and people who picked both got 1000. AKA, people who honestly believed Box B had a million dollars, had Box B get a million dollars. Those who were skeptical, or were otherwise greedy, only got 1000. Therefore, it would seem the best scenerio would be just go for Box B, right?

But what if Alpha thinks you're going to go with both, with whatever voodoo knowledge it uses to determine whether or not Box B has the million or not, because you're usually the type of guy or gal who tends to be adverse to gambling (like me). However, upon reading this thread, you see "well crap, gambling has a 100% proven track record, of course I'm going to go with Box B only and get my million". But Alpha couldn't predict your change of heart, and thus you're left with nothing?

I mean, the opposite could happen to. You could be convinced going for box b is the right choice, because the million is guaranteed under observed conditions. So Alpha generates the million dollar box and leaves. But then you decide after the boxes are generated "screw it, let's get greedy", and pick both boxes. I mean, unless we're dealing with some Schrodinger cat-level metaphyiscs, there is a point where the boxes are generated, Alpha flies off, and you make their choice. There's nothing Alpha can do to change anything once it flies away.

To be frank, I don't know what I'd do. I really would like that million dollars, but I am a pessimistic person by nature when it comes to my own luck, so then I believe that box B will just so happen to be empty if I go with Box B, since the universe hates me. Therefore, to minimize loss, I go with both for the guaranteed 1000. But everyone who genunially believed Box B had the million, the box did contain the one million. I don't know if the people who picked both knew enough of the scenario to believe box B was empty or not. After all, the prompt says that Alpha explains nothing to the player, it just generates the boxes and flies away. We could be unique in the fact that we're the only ones who actually knows the rules of the game before playing the game. as some third party explained to us what the hell was going on.

Ultimately, as I ramble on and probably don't make any sense, its impossible to make a definitive "this is the smarter option" statement since we don't know exactly what causes Alpha to generate the contents of Box B. All we know is that its generated before the game is played, and that Alpha has always been right. But that's gambler's fallacy right there; a coin can land heads 100 times in a row, but that doesn't mean the 101st flip will be heads. Because of this uncertainty inherent in the system, I'd still probably go with both, since I'd rather have a 1000 than nothing. But who knows when I'm actually in the heat of the moment: I really freaking want that million, and I know those who chose a million only, only were the ones who got the million

 

[warpus]

In contrast with religious beliefs, "omega" is providing evidence. That evidence is crushing and consistent. We're not told how omega is doing it, and for the purposes of making a decision it doesn't matter if we hold that it behaves consistently on our choice.

Yeah, he provides statistical evidence that shows that the game is rigged.

 

[TheMeInTeam]

What I will ask you directly is do you deny that Newcomb's problem/paradox, whatever, has legitimately arguably religious/theological implications?

Yes. I deny that. I see no reason to assign credibility or even privilege consideration to beliefs without evidence to a hypothetical situation where key facts to the scenario are dependent on fairly overwhelming evidence.

"Factually inaccurate"? That's a strong claim. Show me where in the Newcomb problem scenario is anything necessarily religious.

I could counter your argument by saying that you're factually inaccurate if you deny that Newcomb's problem is about fishing, since someone out there would use his million dollars on fishing! It has similar rational support.

I mean I understand the position that it can be looked at from a purely mathematical/statistical/probability standpoint... but you seem to be saying it must be, which I disagree with.

Then why not look at it from an upside down goat perspective or an infrared perspective? It is a question of what a decision-maker should do in that scenario to optimize their outcome. Looking at it from a "religious" perspective has identical utility to those two, so why single it out? Maybe you'd get the right answer if you just took an upside down goat perspective.

And TBH, even then, that requires some pretty big assumptions. For one thing... "Omega has been right 100 out of 100 times!" Says who? Omega? Or some guy purportedly speaking on Omega's behalf? And regardless of who said it do you believe it? Why? Did you observe all 100 people getting $1 million with your own eyes, or is this just "something that has been reported." In other words... are you taking it on faith? Why? Because its what you want to hear? Because you've already emotionally committed to the decision?

Now we're getting into objective nonsense. This is a hypothetical scenario. If you don't accept the premise, then the facts involved are irrelevant. I would be very skeptical of a real-life presentation of the Newcomb problem, the person presenting it, and of Omega. Perhaps less if Omega were an alien the size of a large city and I physically observed the boxes dropped in front of me, and much more if another person is telling me about it. For the purposes of doing the problem, I imagined observing these things directly, including 100 accurate predictive outcomes in a row.

To do this thought experiment at all, dismissing it before you begin is not participating in it. Do I "believe" in a hypothetical? That's a wrong question. If you don't accept the hypothetical, there's no reason to consider any of this.

I am not asserting that it can only be looked at in religious terms. I am responding to TMIT's assertion that it cannot be discussed/imagined/framed etc in religious terms and to do so is erroneous and various other adjectives that basically mean "wrong".

Privileging one perspective of hundreds-to-more other perspectives with no reason to do so is a pretty good indicator of bias.

You are assuming that Omega has indeed been infallible for the purposes of the question. There are many other assumptions, you have to make, that is just one. The clergy tells you god is infallible and has always been so. If you are a believer,

Picking the second box is taking a leap of faith as to Omega's prior and continued infallibility.

Now you're changing the problem itself to try to make your point .

Actually, no. Acting on 100 observations is the opposite of a "leap of faith". Faith involves belief without evidence, and the existence of that evidence is a key part of the hypothetical scenario. Newcomb's problem in OP doesn't say "someone told you he saw Omega get it right 100x in a row". The language is "Omega has been correct on each of 100 observed occasions so far". Why ignore that and change the question to suit your need to make it into an upside-down g-...er...religious discussion?

If this were real life and something were presented to you in that way you'd be right to question it, but given that this isn't real life and is in fact a logic puzzle it doesn't really make sense to do so. The whole point of these sorts of logical thought experiments is that you take the premises as read and then analyse what logically follows from them.

+1. This thread is about Newcomb's problem, not a modified version of it with a vague/inconsistent premise.

Yeah, he provides statistical evidence that shows that the game is rigged.

Quite obviously rigged at that. The question is how it is being rigged.

But for the purposes of picking the box that gives you more money, given the information presented, the way it is rigged doesn't matter.

I find it interesting that people immediately jump to breaking causality, "time paradoxes", "god", "faith", and other crap like that. Maybe Omega's simply leaves the boxes based on past observed behaviors, and makes its predictions on factors it has observed most consistently indicate future choices. Maybe Omega is cheating and using technology to simply remove the money after the fact despite apparently leaving, using a trick the observer doesn't notice. Maybe it can run brain scans and use those to make its perfect (so far) predictions.

We don't know. So why assign higher consideration to less likely conclusions? Is a "time paradox" or "god" really the most likely method being applied here? I'd assign a higher probability to "omega" instructing a 3rd party to yank the million dollars through a hidden hole in the floor if someone touches the other box first (though if box A is really disappearing in a puff of smoke, that's interesting too).

Even if someone is actually just pulling that money away from me using a string (a more likely method of rigging this problem), I'd still rather have the money.