Le 22-juin-05, à 20:35, George Levy a écrit :
> Bruno Marchal wrote:
>
>
> Le 21-juin-05, à 05:33, George Levy a écrit :
>
> Note that according to this definition the set of observer states may
> also encompass states with inconsistent histories as long as they are
> indistinguishable.
> The possibilities of observer moment being partially associated with
> (slightly) inconsistent histories resolves the question of how valid
> but erroneous observer moments can exist. For example I could make an
> arithmetical mistake such as 8*5 = 56 or I temporarily believe that
> Christopher Columbus discovered America in 1592.
I agree.
>
> >An interesting thought is that a psychological first person can surf
> simultaneously through a large number of physical OMs
> With comp, we should say that the first person MUST surf
> simultaneously through an INFINITY of third person OMs.
> I agree there is and infinity of OM's that a psychological first
> person surfs through. But I would not say these OM's are "third
> person," because there is no third person to observe them. A
> psychological "third person" would be too spread out among OM's to
> observe any one in particular.
I agree.
>
> (I would not use the term "physical" at all, because at this stage it
> is not defined. But with the negation of comp + assumption of slightly
> incorrect QM what you say seems to me plausible.)
>
> Are you saying that COMP does not admit (slightly) inconsistent
> histories?
No. Quite the contrary, comp does admit inconsistent histories. But for
reason of methodological simplicity, I limit my "interview of lobian
machines" only on the consistent machines, for which comp makes
necessary the consistency of inconsistent histories. With comp, a
consistent machine is in the state of being *possibly* inconsistent.
I know you read the Smullyan's FU, and I'm afraid it is not enough. I
mean it is a non trivial consequence of the incompleteness phenomenon
that a consistent machine is automatically consistently inconsistent,
and this in the frame of the brave and simplest logic (classical
logic).
> I am not sure if I agree with this. I can be a psychological first
> person and still say "yes doctor" to a computer transplant into my
> brain.
Not only I agree, but the point is that with comp it is necessarily so.
Bruno
http://iridia.ulb.ac.be/~marchal/
Received on Fri Jun 24 2005 - 09:44:04 PDT