Re: Data and meaning

Rafael,
Let me try to answer your comments

> Christoph
>
>
> I am not sure if we are thinking in the
> same direction
>
> > Rafael,
> > You higlight a good point.
> > Overall answer: yes the presence of water will generate
> > a meaningful information for you and for the robot.
>
> well only in case the robot is of the kind of being
> that can *experience* (in its own body...) thirst
> (not just to have been *programmed* to *answer*
> 'Yes, I am thirsty' when it perceives water...
>
This thirst is about meaningful information generation as
per MGS structure for a biological constraint, not more.
Obviously, a robot is not a human and there are many
differences beyond this meaning generation. A much
wider horizon is addressed for humans(see hereunder).
>
>
> > But beyond this statement, many things will be different.
> > Presence of water is the external information that you,
> > thirsty human, receive. And this received information will
> > generate a meaning in your body because it has some
> > relation with your constraint "maintain a given level of
> > water in the body". And the meaningful information
> > generated will be "presence of water that can reduce my
> > thirst".
> > For the robot, assuming it has been programmed with a
> > constraint "drink some water", the presence of water will
> > generate a meaningful information "presence of water that
> > can satisfy the constraint".
>
> this is the point where I do not agree: to have been
> programmed is not the same as to *feel* the need
> for water based on bodily constraints... (this is *just*
> a formal similarity). This is the reason, I believe, why
> the *top down* idea of making robots *think* (GPS and
> the like) is *wrong* i.e. it will remain *forever* abstract.
>
Yes, to "feel the need for water based on bodily
constraints" is more than a programmed reaction to a
given parameter. This is why the wording for the
constraint on human in this example has been restricted
to "maintain a given level of water in the body". Which
is a biological constraint in the living body. Very close
to basic life constraints.The comparison is on this basis
only.
Regarding the top down idea of making robots think,
I agree (partly) with you that it will remain "abstract"
as long as we are not in a position to understand what
"think" means. But I cannot say that this "Hard Problem"
will remain a problem forever.
>
> > So the robot, like you, will generate internally a meaningful
> > information. At first look, theses two meaningful information
> > are quite similar. But this similarity is only for this precise
> > meaning vs this precise constraint. For the robot, it is all that
> > is to be said. But for us humans, the presence of water is going
> > to generate many other meanings coming from the constraints
> > related to our history of relations with water and with the world,
> > to our other desires, to our emotions, to our free will, and so on.
> > And for us, the meaning quite similar to the one generated
> > within the robot will be only a little element in the web of
> > meanings that the presence of water will generate within
> > ourselves.
>
> indeed, this is just a *similarity* , here I agree again...
>
> > This makes the difference. As you write it, a human can say
> > "it is related to my whole bodily being-in-the-world with others.. "
>
> > We could also add the other difference about derived vs non
> > derived meaning (the meaning in the robot comes from a
> > human programing. For us, it does not).
> > Apart from this last difference, I feel we can say that it is more
> > a question about a difference of complexity than about a
> > difference of nature.
> > Would you agree ?
> >
>
> yes, but the difference does not just come from the
> progammer, it comes basically from the kind of being
> a robot is (in case it is not a *flesh* one... at least...
>
Yes, and my wording is not good. The kind of being of a
robot is very different from the being of a human. In the
sentence on "difference of complexity", I was implicitly
considering the possibility that in some future the constraints
of human will be understood well enough so they can be part
of a defined MGS. We would then be in a position to talk
about a same general process for meaning generation in human
and in robots. But we are today far from this level of
understanding of human mind, so the sentence should be modified
to "we can say that it could be more a question about ...".
By the way, this brings up a general comment on the
applicability of the MGS to humans: the MGS is a system
based on a constraint to be satisfied. A meaningful
information is generated to satisfy this constraint.
If we do not know the constraint of the system, we cannot
know what type of meaningful information can be generated,
and it is not possible to define an MGS.
Simple cases like basic life have a clear constraint:
survive. Human system is much more complex and partly
unknown. It has many constraint, some we do not understand.
Consequently, we should today limit the usage of MGS for
human to very simple cases (i.e. Closer to basic life than
to performances of human).
Perhaps this point has not been enough explicited in the
FIS 2002 paper "Information and meaning".

Christophe
>
> Rafael
>
> > Regards
> >
> > Christophe
> >
> > > Christoph,
> > >
> > > as far as I can see, something
> > > has meaning for me (or for another
> > > kind of living system) not just if data
> > > are *processed* in my brain but
> > > if it is related to my whole bodily
> > > being-in-the-world with others... I mean,
> > > if I am thirsty it is because my body is
> > > of this kind that it needs water. Would
> > > it make *any difference* (i.e. any
> > > meaning and therefore has any
> > > informational value...) if a robot that
> > > is not of the kind of bodily existence
> > > as I am, would be able to say: I am
> > > thirsty? and the same with all other
> > > *meanings*.
> > >
> > > Rafael
> > >
> > >
> > >
> >
> >
>
>

Regards
Christophe Menant (crmenant@free.fr)
Received on Thu Oct 3 09:23:39 2002