Re: Reality, the bogus nature of the Turing test

From: Bruno Marchal <marchal.domain.name.hidden>
Date: Mon, 18 Sep 2006 14:46:48 +0200

Le 16-sept.-06, 23:37, Colin Geoffrey Hales a crit :

>
> Bruno wrote
>> Colin Geoffrey Hales a it :
>>> 5) Re a fatal test for the Turing machine? Give it exquisite novelty
>>> by
> asking it to do science on an unknown area of the natural world. Proper
> science. It will fail because it does not know there is an outside
> world.
>>
>>
>> And you *know* that?
>>
>> We can *bet* on a independent reality, that's all. Justifiably so
> assuming comp, but I think you don't.
>>
>> Self-referentially correct machine can *only* bet on their
>> self-referential and referential correctness.
>>
>>
>> Bruno
>
> I don't assume COMP. The idea that this is necessary to hold a
> position on
> anything is, for me, simply irrelevant and preumptuous that COMP is
> able
> to make any useful predictions.


My point is that COMP is a so big assumption that it does make
verifiable predictions.




> COMP is not an empirically supportable
> position, no matter how elegant it may look.

But is there any empirically reason to disbelieve in it?
 From the UDA you can show that to make comp false you need to introduce
actual infinities in the subject.
I know only Penrose for having try to do that, unsuccessfully.




> I would consider it so if it
> could predict the existence and properties of brain material.

But that is exactly my modest (UDA) point. Comp has to predict the
(apparent) existence of the brain, atoms ...
I show why.
Then I show how and got results in the arithmetical UDA (or lobian
interview).



>
> Having said that ....yes you are right that 'betting' on an independent
> reality is all we can do....this is an empirical matter.

All right. Note that if you believe in primitive physicality, you are
quite coherent by abandoning comp.



> Whatever it is
> that enables vast legions of scientists to do their job (deines their
> job), relentlessly for hundreds of years....that mutually eqisitely
> produced, shared delusion called the natural world.... that thing...
> that
> we appear to be within and constantly demonstrate it via creation of
> novel
> technology that seems to operate within it.... That is worth betting
> on...the process of consideration that it may not be there is of no
> practical value.

I do agree with you, but let us not confuse the two following bets:
a) Betting on an independent reality (like I do)
b) Betting on a material primitive world (like I do not).

I have always feel myself as a REALIST scientist. But then I argue that
if comp is true, then physical stuff emerge from a deeper non material
reality, like for example (assuming comp) the relation between numbers.

Perhaps even Stephen Hawking points in that direction with his
beautiful selected basic papers: "God created the Integers".


>
> But I'm not sure you have really 'got' what I mean by 'it does not know
> there is an outiside world'. This is a practical matter. Brain material
> does something special...which enables an internal literal phenomenal
> mapping of the universe outside the scientist. The Turing machine is a
> collection of abstractions with an ASSUMED relationship to the outside
> world.

Anything talking about anything supposedly outside itself makes such
an assumption.



> Until we know what that physics is any argument assuming the lack
> of that special physics is simply going to take you down the usual
> argument path of assumption.
>
> Only when we isolate the real physics of phenomenal consciousness in
> brain
> material can we then make any valid judgement as to its necessity in
> intelligence. Until then I hole all discussion based on assumption of
> computational (as-if) substrates as invalid or at least interesting
> but of
> little practical use at this stage.


No problem.



> -------------------------------------------
> TURING TEST.
> The turing test always infuriates me. Since when does dumbing a human
> down
> to the point of looking like machine X prove that machine X has
> consciousness? I just don't get it.


Turing was really not searching any proof there.



>
> When you give the machine that faculties of a human and make it do what
> humans do ...I I believe getting them both to do science is the
> appropriate ttest... then the Turing test is a complete irrelevance
> based
> on an assumption that the presence of the physics of phenomenal
> consciousness is optional in intelligence. It is an empirical reality
> that
> when you alter phenomenal consciousness then scientific behaviour is
> altered. No further argument is needed. The turing test is not a test
> of
> consciousness. I'm not sure what it is a test of, but it is certainly
> not
> a test of consciousness.


I think Turing would agree here, except that it would have add that
such a test is the better thing you can ever have to evaluate the
plausibility of the presence of consciousness (without being influenced
by the prejudices based on body shapes). I am less sure because with
the technical progress only arbitrary longer test can make sense. I
know someone who did took some program for a conscious being after a
short "conversation" with it!

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.domain.name.hidden
To unsubscribe from this group, send email to everything-list-unsubscribe.domain.name.hidden
For more options, visit this group at http://groups.google.com/group/everything-list
-~----------~----~----~----~------~----~------~--~---
Received on Mon Sep 18 2006 - 08:47:34 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:12 PST