This is the standard argument for one-boxing. The two-boxer will reply, “But the boxes are already filled!” The one-boxer replies “One boxing wins!” The two-boxer replies “THE BOXES ARE ALREADY FILLLLED!!!” The one-boxer replies “BUT. ONE. BOX. ING. WINNNNS!!!”
A paradox is not resolved by clinging to one side of it and claiming it refutes the other.
Here is another variation on the problem. Suppose you discover how Omega makes its predictions. It turns out that there is a gene whose different alleles predispose you to one-boxing or two-boxing on Newcomb’s problem. (Hey, this is no sillier an idea than in a lot of thought experiments.) If you have variant 1, then 99% of the time you will one-box, and similarly for variant 2. Omega is, in effect, telling you with 100% reliability which variant you have, and has filled the boxes accordingly.
No-one has previously faced Omega with that knowledge. What do you choose?
This is the standard argument for one-boxing. The two-boxer will reply, “But the boxes are already filled!” The one-boxer replies “One boxing wins!” The two-boxer replies “THE BOXES ARE ALREADY FILLLLED!!!” The one-boxer replies “BUT. ONE. BOX. ING. WINNNNS!!!”
A paradox is not resolved by clinging to one side of it and claiming it refutes the other.
This video may be illuminating: Ilya Shpitser’s talk on Newcomb’s problem at FHI.
Here is another variation on the problem. Suppose you discover how Omega makes its predictions. It turns out that there is a gene whose different alleles predispose you to one-boxing or two-boxing on Newcomb’s problem. (Hey, this is no sillier an idea than in a lot of thought experiments.) If you have variant 1, then 99% of the time you will one-box, and similarly for variant 2. Omega is, in effect, telling you with 100% reliability which variant you have, and has filled the boxes accordingly.
No-one has previously faced Omega with that knowledge. What do you choose?