Box B is already empty or already full [and will remain the same after I’ve picked it]
Do I have to believe that statement is completely and utterly true for this to be a meaningful exercise? It seems to me that I should treat that as dubious.
It seems to me that Omega is achieving a high rate of success by some unknown good method. If I believe Omega’s method is a hard-to-detect remote-controlled money vaporisation process then clearly I should one-box.
A super intelligence has many ways to get the results it wants.
I am inclined to think that I don’t know the mechanism with sufficient certainty that I should reason myself into two-boxing against the evidence to date.
Does it matter which undetectable unbelievable process Omega is using for me to pick my strategy? I don’t think it does—I have to acknowledge that I’m out of my depth with this alien and arguments against causality defiance or the impossibility of undetectable money vaporisers are not going to help me take the million.
Another tack: Omega isn’t a super intelligence—he’s got a ship, a plan, and a lot of time on his hands. He turns up on millions of worlds to play this game. His guesses are pretty lousy, he guesses right only x percent of the time. We are the only planet on which he’s consistently guessed right. We don’t know what x is in the full sample size. Looking at what his results are here, it looks good. Does it really seem rational to second guess the sample we see?
It seems to me that we have to accept some pretty wild statements and then start reasoning based on them for us to come to a losing strategy. If we doubt the premises to some degree then does it become clear that the most reasonable strategy is one-boxing?
Box B is already empty or already full [and will remain the same after I’ve picked it]
Do I have to believe that statement is completely and utterly true for this to be a meaningful exercise? It seems to me that I should treat that as dubious.
It seems to me that Omega is achieving a high rate of success by some unknown good method. If I believe Omega’s method is a hard-to-detect remote-controlled money vaporisation process then clearly I should one-box.
A super intelligence has many ways to get the results it wants.
I am inclined to think that I don’t know the mechanism with sufficient certainty that I should reason myself into two-boxing against the evidence to date.
Does it matter which undetectable unbelievable process Omega is using for me to pick my strategy? I don’t think it does—I have to acknowledge that I’m out of my depth with this alien and arguments against causality defiance or the impossibility of undetectable money vaporisers are not going to help me take the million.
Another tack: Omega isn’t a super intelligence—he’s got a ship, a plan, and a lot of time on his hands. He turns up on millions of worlds to play this game. His guesses are pretty lousy, he guesses right only x percent of the time. We are the only planet on which he’s consistently guessed right. We don’t know what x is in the full sample size. Looking at what his results are here, it looks good. Does it really seem rational to second guess the sample we see?
It seems to me that we have to accept some pretty wild statements and then start reasoning based on them for us to come to a losing strategy. If we doubt the premises to some degree then does it become clear that the most reasonable strategy is one-boxing?