I don’t think this argument holds up. Suppose the holodeck hypothesis is true and someone outside the simulation decides to punish irrational choices by killing you if you buy a lottery ticket. The probability of you being killed is around 10^-3 so you should never risk buying a lottery ticket.
The problem is that you’ve no reasonable basis for assigning your 10^-3 probability for a good outcome rather than a bad outcome or a batshit insane outcome. You also have no basis for your 10^-4 probability of being in a holodeck. The only rational way to behave is to act as if you’re not in a holodeck (or a world with an occasional interventionist god who does his damndest not to ever leave clear proof of his interventions, or a simulation run by aliens, or The Matrix) because you have no basis on which to assign probabilities otherwise. This changes of course if you are confronted with evidence that implies a greater likelihood of one your holodeck hypothesis.
The problem is that you’ve no reasonable basis for assigning your 10^-3 probability for a good outcome rather than a bad outcome or a batshit insane outcome. You also have no basis for your 10^-4 probability of being in a holodeck. The only rational way to behave is to act as if you’re not in a holodeck (or a world with an occasional interventionist god who does his damndest not to ever leave clear proof of his interventions, or a simulation run by aliens, or The Matrix) because you have no basis on which to assign probabilities otherwise. This changes of course if you are confronted with evidence that implies a greater likelihood of one your holodeck hypothesis.
OK, what if we accept the simulation hypothesis, and we further say that the advanced civilizations are simulating their ancestors? Then we’d expect our simulators to be evolved or derived from humans in some way. It’s unlikely we’ll change our ideas of fun and entertainment too much, as those are terminal values—we value them for their own sake. This gives us some pretty strong priors, based on just what we know of current human players...
I’m not sure how to interpret that though. Based on players of the Sims, say, we would expect tons of either sadistic deaths by fire in bathrooms and starvation in the kitchen, or people with perfectly lovely lives and great material success. Since we don’t observe many of the former, that’d suggest that if we’re in a simulation, it’s not being run by human-like entities who intervene.
I don’t think this argument holds up. Suppose the holodeck hypothesis is true and someone outside the simulation decides to punish irrational choices by killing you if you buy a lottery ticket. The probability of you being killed is around 10^-3 so you should never risk buying a lottery ticket.
The problem is that you’ve no reasonable basis for assigning your 10^-3 probability for a good outcome rather than a bad outcome or a batshit insane outcome. You also have no basis for your 10^-4 probability of being in a holodeck. The only rational way to behave is to act as if you’re not in a holodeck (or a world with an occasional interventionist god who does his damndest not to ever leave clear proof of his interventions, or a simulation run by aliens, or The Matrix) because you have no basis on which to assign probabilities otherwise. This changes of course if you are confronted with evidence that implies a greater likelihood of one your holodeck hypothesis.
OK, what if we accept the simulation hypothesis, and we further say that the advanced civilizations are simulating their ancestors? Then we’d expect our simulators to be evolved or derived from humans in some way. It’s unlikely we’ll change our ideas of fun and entertainment too much, as those are terminal values—we value them for their own sake. This gives us some pretty strong priors, based on just what we know of current human players...
I’m not sure how to interpret that though. Based on players of the Sims, say, we would expect tons of either sadistic deaths by fire in bathrooms and starvation in the kitchen, or people with perfectly lovely lives and great material success. Since we don’t observe many of the former, that’d suggest that if we’re in a simulation, it’s not being run by human-like entities who intervene.