A Million Dollar Puzzle: The Newcomb Paradox The Game The game has the following form. You will be presented with two boxes: *Box A will contain $10,000. *Box B will contain EITHER $1,000,000 or nothing. You will not be able to see inside the boxes, so you won't know whether Box B contains $1,000,000 or nothing. You will then be given the choice between taking home both boxes or just Box B. There are two crucial things to take into account when making your decision. *The amount of money in Box B will be determined by a supercomputer on the following basis: If it predicts that you will take home both boxes, then no money will be put into Box B. If it predicts that you will take home only Box B, then $1,000,000 will be placed into the box. The supercomputer has been 100% accurate in its predictions in the past. *The prediction will already have been made by the time the game starts, and the amount of money in Box B is already fixed. Assuming you want to win as much money as possible, should you take home both boxes or just Box B?
It's not really a puzzle if it's a paradox. You could flip a coin to help remove selection bias to trick the computer, but it won't matter. If the computer's prediction rate is 100% rather an extremely high probability (i.e 99.99999~ %) then it could theoretically predict the coin flip and a computer-based random number generator, which is still limited based on preset algorithms in the programming. It also defeats the purpose of game theory and rational utilitarianism if there's zero error margin. The only really solution is to base it on some kind of quantum dynamics and random subatomic particle interactions which the computer wouldn't have the processing power to predict, but again, it becomes moot if the computer is infinitely powerful. Take home both boxes because there's no logical alternative.
What if it's accurately guessed 1 out of 1 attempts? That's 100% accuracy, but doesn't speak to its overall ability over time. Never trust percentages. The whole premise of a "supercomputer being all-knowing and all-powerful" is a non-starter to me. Might as well say "a psychic has 100% success rate" and I'd believe it more.
Either there was a typo here or this is the simplest fucking puzzle in existence. The computer will put a million dollars into Box B if it predicts that you will only take home Box B, and it is always accurate in its predictions. So you fucking take just Box B and collect your million dollars.
No typo, the key is that it assumes you act in your own self-interest and take home the most money possible. The most you can take home is $1,010,000. So by only choosing Box B, you're not doing that, so theoretically you lose. Potential payouts:
Either my reading comprehension REALLY sucks, or this is the easiest question in the world. Answer: You take home both boxes, because best case is you get $1,010,000, worst case is you get $10k. All the mumbo jumbo that I deleted from the quote and the title of "newcomb paradox" is irrelevant, and meant to confuse you into thinking this is a "one or the other" type question because it all sounds so serious.
So this is like Schrodinger's Wallet. The box contains money until you open it and the waveform collapses and it either contains $1,000,000 or $0.
The problem here is that you know that the computer is omniscient. So even if you try to trick it it already knows. Thus the two options where the predicted choice and the actual choice are not the same can be thrown out. The computer knows man. So you'd only take Box B. And the computer knows the whole time you'll take Box B. With the computer being omniscient and these parameters there's no way to get $1,000,010.
I You're assuming the computer is omniscient. It's never said though. I took the assumption that is programmed to assess information about your personality and previous actions to conclude the most likely scenario. I agree that you take both home and be happy with whatever you get. Also it states the computer made the decision first so your decision has no effect on the outcome.
Option C: Kick that computer in the hardware, pull a gun on Howie Mandel, and run off with the full $1.01 million. I naturally assume Howie Mandel is involved in something this stupid. And it being televised on FOX. A better conundrum would be to do a game based on The Pit And The Pendulum where you have to choose to save your girlfriend or a pack of cold beer, but you don't know what type of beer it is. RATINGS GOLD.
For me personally, disregarding any attempts to outsmart the computer or whatever, I would always take Box B (and only Box B) home. $10,000 is not a life-changing amount of money, and $1,010,000 is not meaningfully more than $1,000,000. So a shot at $1 mil outweighs the guarantee of taking home some amount of money. If I were broke enough that I could change my world with $10k, maybe I'd go with the guarantee.
That's why I say, if the computer is the controlling factor, hack the fucking computer and rig the outcome in your favour. You're welcome.
I remember reading this when I was ten in Martin Gardner's "Aha! Gotcha" book on paradoxes. There are two thought processes in here: 1: (Take both boxes) The amount of money in B is not going to change. It's either there or it isn't. It's already made its decision, after all. So, the fact that I'm going to decide to take both boxes isn't going to influence it at all. If it's decided to screw me, I have no control over that. I might as well get some money out of the deal. 2: (Take one box) Well, the computer is simulating your thought process and knows which decision you're going to come to. It's been 100% accurate in the past, and its creator is pretty confident that it's going to be 100% accurate with you. So it's going to see that conflict in your mind, determine that you display Mindset #1, and give you an Empty Box B. So, it's better to determine that you're going to do Mindset 2 and just take Box B. The computer, being omniscient, will predict this and give you the million. I find myself sympathizing with Thought Process 2 here. If the computer is completely omniscient, then the correct choice is to take just Box B. Amusingly enough, the best way to get the million is to be a retard who goes "A million is more than a thousand. I'm gonna take Box B." By not appreciating the subtlety of the puzzle, it's guaranteed that he's going to get the million.
Why not make the puzzle easier and say that box B has a 50/50 chance that you will get a million? And you can only pick one. A guarantee that you get 10k, or a chance to go bigger? And if you do take the chance, how low of a chance are you willing to accept? 40%? 25%? If this computer is already predicting what you will do, then you have to be stupid not to take just box b, especially if you know the background behind the game and how it comes to a conclusion. The object of the puzzle is to come away with the most money, but it's flawed in that 1.01mil is only 1% greater than the 1mil. Any logical person would chose the guaranteed mil.
Don't over think it. Take both. You're guaranteed 10,000 and might get 100,010,000. Who wouldn't benefit from 10K?
I still maintain my original answer is the correct one. Namely, that it's a trick/obvious question. Because I thought I was massively missing the point at first, I researched this, and it's astounding how many "academics" (read: collegiate lifers who are afraid of the real world) think this question is a legitimate thing. And it seems like, the more you over-think something, the more you miss the point. It's annoying how many people make a living off annoying others.
A lot of academic problems sound really, really stupid. The problem is that the answers to these questions often touch on ideas that are fundamental to the discipline, like free will and determinism in this case. These guys aren't really debating this question by itself; they're debating the implications of saying whether one answer is better than the other one. "Well, if we say that taking both boxes is the right choice, it has the following implications... and if we say that taking Box B is the right choice, then it has another set of implications. And if we say that there's no good answer, it has a third set of implications. Which one do we want?" Cue a massive debate that looks like a neckbeard argument over Star Trek characters to everyone else. It doesn't help that in many cases, academics boil down their century-old debates to questions like this and don't say what the history is behind it.