In coherence arguments, we deal with many alternative situations, or thought experiments, asking what is to be done in each of them, considered separately from all the other situations. Presenting any one situation can be taken as a form of updating from some shared prior state on the situation’s description, or on an observation of it, or on the newly presented state of knowledge of it. This is different from ordinary updating that follows a worldline and proceeds in order; instead, the updates that put the point of view in one of these situations happen independently of each other.
In this framing, an anthropic update on the birth rank makes sense, even as it would be much less central as an ordinary observation. It can successfully fail to mention anything else that an actual human would otherwise know, so that the resulting state of mind will have large gaps in its knowledge of the world. The point of the procedure is to go from a statement of what the birth rank is to some posterior state of mind that describes the situation, captures the given facts, but no more facts than that.
And clearly this is a strange place to stop, without going further to elicit a utility function, or to find an updateless strategy. Yet that’s what anthropics tends to be about, building up a context that’s sensitive to technical details that can go unmentioned, motivated by conflicting desiderata from other pursuits that are considered out of scope, and leaving it at that.
In coherence arguments, we deal with many alternative situations, or thought experiments, asking what is to be done in each of them, considered separately from all the other situations. Presenting any one situation can be taken as a form of updating from some shared prior state on the situation’s description, or on an observation of it, or on the newly presented state of knowledge of it. This is different from ordinary updating that follows a worldline and proceeds in order; instead, the updates that put the point of view in one of these situations happen independently of each other.
In this framing, an anthropic update on the birth rank makes sense, even as it would be much less central as an ordinary observation. It can successfully fail to mention anything else that an actual human would otherwise know, so that the resulting state of mind will have large gaps in its knowledge of the world. The point of the procedure is to go from a statement of what the birth rank is to some posterior state of mind that describes the situation, captures the given facts, but no more facts than that.
And clearly this is a strange place to stop, without going further to elicit a utility function, or to find an updateless strategy. Yet that’s what anthropics tends to be about, building up a context that’s sensitive to technical details that can go unmentioned, motivated by conflicting desiderata from other pursuits that are considered out of scope, and leaving it at that.