Russell wrote:

> The empirical problem with the ASSA is that under most reasonable
> proposals for the absolute measure, observer moments corresponding to
> younger people have higher measure than older people. Whilst the
> reference class issue puts a lower bound on how old you would expect
> to be, it seems unlikely that children aged 4 or 5 could be considered
> excluded from the reference class - I can remeber being conscious at
> that age, and children of that age seem conscious from the outside.

I don't understand why measure should decrease with the age of a person. Of
course, once you take into account the possibility of dying then you will
see a decrease. But ignoring that, the measure should be conserved. The
measure for being in a particular state at age 30 should be much smaller
than the measure for being in a particular state at age 4, but after
summation over all possible states you can be in, you should find that the
total measure is conserved.

> The second problem with the ASSA is lack of subjective time. I have
> always argued that subjective time is necessary to experience anything
> at all. This is a direct consequence of computationalism, but I think
> is more basic than computationalism (since I don't really count myself
> as a computationalist).

Let's define observer moment as a program in some computational state (maybe
it is better to consider programs in different computational states as
distinct programs). That program must have information stored in itself
about past events and must have a sense of a subjective time.

> Now I don't expect to convince you of this - I never succeeded in
> persuading Jacques Mallah. However I do want to point out that even
> with the ASSA, one should not expect to experience survival of the WTC
> with some loss of memories. The most likely outcome is another
> observer moment with high measure - namely being a newborn baby. In
> other words, what you'd experience is reincarnation.

I think that if you take into account the ''entropic'' effect of there being
more states for you to be in at higher age, then you can end up at a large
range of ages.

> To get the effect you were suggesting would require another type of
> SSA, about which I have complete failure of imagination.

I think it is similar. You have a set of all universes which we identify
with descriptions or programs. Embedded in these descriptions are
descriptions of self aware substructures. A measure on the set of all
programs defines a measure on the set of all substructures. I then say:
''That's all there is''. The proponents of RSSA go further and postulate new
rules about what the next experience of a SAS should be. What you are
actually doing is promoting our experience of the flowing of time to
fundamental law. However, this is something that should be derived from more
fundamental concepts.


Reply via email to