I don’t know whether I am Alice or SimAlice, and as there is one of each, there is a 0.5 probability of me being either one
I’m having a hard time justifying the 50% here. I don’t buy it at all, actually. But that does’t mean SCDT shouldn’t use it: if we take it to be proportional to the number of known simulations, we run into a lot of trouble (let’s call this #SCDT)
Omega: In transparent box 1, here’s $1000. You also get what’s in the opaque box, which is 1 cent if I predicted you would choose to only take the opaque box, otherwise it’s empty. I determine this by simulating you one trillion times.
what #SCDT says:
1 real Alice: take the transparent box for $1k
1 trillion simulated Alices: if we take only the opaque box, real Alice will get 1 more cent! $0.01
Net outcome: Alice walks away with 1 cent.
Meanwhile, SCDT has Alice walks away with $1k.
I suspect that by weighting them equally, you’re simply allowing yourself to find whichever option works out better on the net. The 50% probability is specious.
I’m having a hard time justifying the 50% here. I don’t buy it at all, actually. But that does’t mean SCDT shouldn’t use it: if we take it to be proportional to the number of known simulations, we run into a lot of trouble (let’s call this #SCDT)
Omega: In transparent box 1, here’s $1000. You also get what’s in the opaque box, which is 1 cent if I predicted you would choose to only take the opaque box, otherwise it’s empty. I determine this by simulating you one trillion times.
what #SCDT says:
1 real Alice: take the transparent box for $1k
1 trillion simulated Alices: if we take only the opaque box, real Alice will get 1 more cent! $0.01
Net outcome: Alice walks away with 1 cent.
Meanwhile, SCDT has Alice walks away with $1k.
I suspect that by weighting them equally, you’re simply allowing yourself to find whichever option works out better on the net. The 50% probability is specious.