Newcomb’s Paradox

The Ask Marilyn thread debate reminded me of my favorite problem from decision theory. It’s called Newcomb’s Paradox though Robert Nozick is really the person that made it famous.* It tends to cause as many insistent split opinions as the Monty Hall problem does.



Anyway, it goes like this.

There is a famous predictor of human behavior. He can predict human behavior with near certainty. Maybe it’s based on new brain scanning technology or maybe he’s just the greatest poker player to have ever lived, but in any event, this predictor knows with near certainty exactly what you (or anyone else) is going to do. Why this predictor decided to run weird experiments on humanity rather than make millions in the stock market is beyond the scope of this problem.

The predictor calls you into a room and asks you to sit down. He tells you that after studying your behavior for several weeks he has made a prediction about what you will do.

In front of you are two boxes A and B. Inside Box A the predictor put $100. Inside Box B the predictor put either:

1) $0 if he predicted you would take both boxes A and B; or

2) $1,000,000 if he predicted you would take only Box B.

So the question in the poll, is what do you do? Take both boxes or just box B.

I'm going to close this thread. Please pick it up here: THE NEWCOMB PROBLEM.

btw, I think I've changed my answer on this since creating that thread.