On 01 Feb 2011, at 07:51, Colin Hales wrote:
Hi Bruno,
I have been pondering this issue a bit and I am intrigued about how
you regard the problem space we inhabit. When you say things like ...
"Are you aware that If comp is true, that is if I am a machine ..."
I cannot fathom how you ever get to this point.
By looking at amoeabs, then reading book on molecular genetics,
smelling Turing universality, then by reading Gödel's proof and the
discovery of how to handle self-duplication and self-reference in
representational machine, ...
I did not take this too much seriously until my understanding of
Church thesis deepens. The closure of computerland for diagonalization
makes universal machine extremely universal, if I can say.
This is a presupposition that arises somehow in the lexicon you have
established within your overall framework of thinking.
It has lead me to some interest with that hypothesis.
Let me have a stab at how my view and yours correlate.
In my view
========================================================
A) There is a natural world.
We, Turing machines dogs, computers are all being 'computed' by it.
This is a set of unknown naturally occurring symbols
The natural 'symbols' interact naturally.
This is 'natural computation'. NOT like desktop computing.
Universe U ensues.
Scientist S is being computed within U
Scientist S can observe U from within.
U makes use of fundamental properties of the symbols to enable
.... observation, from within. Call this principle P-O
If by natural world you mean the world of the natural numbers with
addition and multiplication, I am OK. I can picture your "A)".
If by natural world you mean the physical worlds as seen by 'numbers',
what you say might be locally correct, but that remains to be proved
(assuming comp).
B) This is a symbolic description of U created by S from within U
S can concoct a description of the natural symbols in (A)
It need not be unique, many (B) correspond to one (A)
S can never know if it's completely done.
S can never know the real nature of the sybols in (A)
Descriptions (B), with P-O, explains observation and the observer S
C) There is a _second_ description
It is also concocted by S
These are the normal empirical laws we all know so well
?
It describes how the U appears to S from inside
It need not be unique, many (C) correspond to one (A)
No (C) ever explains observation.
In this framework
(i) a computer running description/rules (B) is not the natural world.
OK. With the two sense of natural world I accept above.
(ii) a computer running description/rules (C) is not the natural
world.
OK.
(iii) a computer running descriptions (B) or (C) is 'artificially
computing'
Yes. it is an isolated malin génie.
(iv) (C) is physics that present day scientists construct
I don't get "C".
(v) (B) is physics of a natural world prior to an observer.
This exist for Löbian machine (although they can find it "looking
inward").
(vi) (A) is 'NATURALLY computing' in the sense that it is literally
'computing' scientist S.
=====================================================
OK.
These options are the logically justifiable position we can take
when we are, as we are, inside U trying to work U out from within,
using an observation faculty provided by U as part of (A). Empirical
evidence justifying (C) is normal overvation (contents of one or
more observer-agreed conscious experisnces). Empirical evidence
justifying (B) is implicit in the existence of an observer
concocting a set (C). You can't be confused about an bservation
unless there is an observer to be confused.
=====================================================
All that said.....now ....
You mention "digital physics". You say "Are you aware that If COMP
is true, that is if I am a machine ..."
In terms of my framework....you are speaking of ...what?
I postulate, eventually, only natural numbers and addition and
multiplication. Then from this (it is not obvious but standard in good
logic textbook) you can show that the arithmetical relation (defined
with "+" and "*", and classical logic) emulate all computations.
Physics or the natural world is never emulated (but often simulated by
malin génie program). Physics is what appear from inside taking the
first person indeterminacy inyto account. A priori the natural world
is not a computational object.
(1) A 'Turing machine (digital computer)' inside U running (B)
descriptions?
(2) The natural computation itself, of kind (A)?
I suspect
(3) Some kind of magical 'computer' in idea-space computing us as (A)?
i.e. A 'virtual machine' that 'acts as if' it generates an arbitrary
number of different U?
The COMP I talk about having refuted is in (i) or (ii) above.
I suspect this is not the COMP you are speaking of...
The comp I talk about is the assumption that my (generailzed) brain
can be emulated by a digital computer. The rest should follow.
As far as I can tell we're not even on the same page. Maybe others
here are in a similar position and don't know it.
I hope you can help.
What do you answer to Stathis Papaoiannou question? By putting special
actual infinities in your body, you might escape the comp hypothesis,
in which case we are indeed not on the same page.
NOTE: When I say I want to build an artificial general intelligence,
I say I can build, within (A), using chunks of (A), an inorganic
observer of kind S, say S', that will also be able to observe and
concoct (B) and (C). S' will NOT be 'artificially computing' rules
(B) or (C)! There will be some symbolic manipulation in the
hardware, but this is not S', it merely drives some of the S'
hardware, like the rules that drive synaptic plasticity. Background
housekeeping. In that event, in my framework, the natural world (A)
will be 'computing' S' too. The properties of (A) called P-O above,
that make S observe also operate within S'. The explanation of HOW
observation happens is in P-O as it is configured in (B).
Is S' conscious?
Best,
Bruno
http://iridia.ulb.ac.be/~marchal/
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to
everything-list+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
