[extropy-chat] Fragmentation of computations
lcorbin at rawbw.com
Tue Apr 3 03:50:17 UTC 2007
> The subjective mode is the important part though, isn't it?
I would say that "the subjective" is indeed important, but perhaps only
to give flavor :-) or perhaps only as an aid to subjective description.
I conjecture that everything that has any real substance may be
reduced to an objective account, and moreover, anything that
cannot be so reduced ought to be looked upon with a great deal
> It presents something of a paradox, because you feel that you are
> living a full life when in fact you are not.
I would counter that if in fact you are not living a full life, then you ipso
facto do not fully feel that you are. In the case, say, where 9 out of
10 of your states are looked up (shelving for the time being the more
complicated scenarios of HashLife), then while you report to everyone
that everything is fine, the "you" that is so reporting is present only to
Now certainly, under some of these extreme conditions we have been
considering (where I contend that functionalism fails), a robot whose
so-called consciousness is not being calculated but merely looked up
does give every appearance of being conscious. Surely he does. But
then, it's also true that sufficiently primitive people will be fooled by
> What about the inverse situation where you are granted an extra second
> of life for every second lived, but remember nothing of that extra second?
> What about all the extra copies of you in branches of the multiverse no
> longer in your potential future? I know that your view is that copies are
> selves and it is just a matter of summing the total runtime.
Yes, that is what I say.
> This is a consistent way out of the paradoxes of personal identity raised
> by the thought experiments with which we are familiar.
> My way is to deny that there is any self persisting through time at all,
> but accept that there is an illusion of this due to the way our brains
> have evolved, and "survival" consists in maintaining this illusion.
Yes, the concept of *self* has come under a lot of fire. But then I
suppose that we agree that what is important is survival. I then take
the admittedly lazy step of saying that in this case, it is a "self" which
survives! Why not? Even if there isn't any such thing on close
examination, as you suggest, it still conveys the idea we are after
(The terms "myself", "self", and so on are pretty thoroughly embedded
in the language, and of course people are going to go on using these
terms. And as for "I" and "me", we'll never get rid of them, and I'm
sure it's not even worth the effort.)
> > And then you write "and no one else can tell the difference...". But I
> > think that that is false. I believe that the scientists observing the
> > phenomenon (a subject getting runtime) determine that there *is* no
> > subject (he doesn't exist as a conscious entity) during those times in
> > which his states are merely being looked up.
> How would the scientists know that there is no consciousness present
> when the behaviour is the same? It is like trying to decide if a robot is
Scientists will conjecture that consciousness is or is not present in the same
way that they make conjectures in other theories. For example, I would
hold the tentative belief, were I introduced to you, that you were conscious.
If, however, a rigorous examination of the way that you were computed
showed that you were only a GLUT (and were not being computed at all
in the nice usage of terms I favor), then I would cease to believe that you
Likewise, we hopefully into the future keep to the principle "conscious entities
should not be made to suffer", at least in the public parlance. We simply
value what people do (and do not) experience. From this, it could easily
follow---provided that the concept does not become obsolete---that from
suitable comparisons with the things we know are not conscious (such as trees
and rocks) and from comparisons with things that we know *are* conscious,
e.g. human beings, the correct physical correlates can be teased out.
> Returning to your example of driving to avoid an accident, imagine you
> are a being in a Life simulation. You come to a point where you can
> either slow down or keep going and run over a pedestrian. You decide
> to slow down, because you think that running people over is bad and
> because you think you have control over your life. In reality, you could
> not do other than slow down: that was determined in the Life universe
> with the force of a mathematical proof. Your feeling that you could have
> acted otherwise is entirely illusory, as you could no more have changed
> what was to happen than you could, through much mental straining, have
> changed 16 into a prime number.
That's right. If we are living in a deterministic universe and something
happens, then we must admit that from an external viewpoint or
according to the physics running our simulation, what happened was
But see Dennett's nice discussion of "evitability" in his book Freedom
Evolves. He considers exactly this case, namely, a conscious Life
entity, and he shows to my satisfaction that there is a strong, meaningful
sense in which one not should regard future hazards as inevitable.
More information about the extropy-chat