Re: subjective reality

From: Bruno Marchal <marchal.domain.name.hidden>
Date: Fri, 26 Aug 2005 17:35:13 +0200

On 24 Aug 2005, at 05:02, Stephen Paul King wrote:

>> BM:
>> Not at all and it is a key point. You confuse what I call comp, (I
>> am a machine, "Yes doctor", en gros), the strong AI thesis, that
>> is machine have phenomenal qualia (say), and BEH-MEC, behavioral
>> mechanism: machine can behave *like* if they had phenomenal qualia.
>>
>
> [SPK]
>
> This looks suspiciously like confusing 1st person and 3rd person
> aspects! But whatever the case, I disagree. Unless a means can be
> found to record and playback (as in the movie Brainstorm)
> phenomenal qualia we have nothing but factually unjustified belief
> in YD. With out proof all that one has is BEC-MEC, i.e. the Turing
> Test.


I don't understand any sentences in that paragraph. If you have the
time to develop it a little bit, I would appreciate.



>>
>> To be clear: to refute mec-beh you need to prove that ALL machine
>> (note one!) fail the turing test (en gros), to refute the strong
>> AI thesis, you need to prove that ALL machine cannot have
>> phenomenal experiences (or subjective, first person, private, etc.).
>>
>
> [SPK]
>
> Not really, all that is required is that it is in principle
> impossible for the class of Machines to emulate minds. IF minds are
> purely classical, AI goes through. IF minds as some aspect that is
> QM that is indispensable, the proof holds. QED.

Remember that comp could be true even in the case my brain is a
quantum computer. That makes the "doctor" and "teleportation" thought
experiments more complex, but the once the Universal Dovetailer is
invoked those supplementary difficulties disappear.


> Unless there exists a subclass of Machines that satisfies the all
> of the requirements to emulate an arbitrary Mind then Strong AI
> is ...


You are unclear for me, I'm afraid.


>>
>> To refute comp (see the definition in my SANE paper) you need to
>> show that for all level of digital description of yourself, none
>> can be turing emulable.
>>
>
> [SPK]
>
> This is too high a bar to ask for! In effect you have made comp
> unfalsifiable!


Not at all. Remember that I have shown that if comp is true then
physics is given by a precise theory. If that theory is in
contradiction with known physical facts then comp is refuted. Comp
has already pass the test of the non booleannity of the observable
propositions.


> Like I stated in the first place, you are asking for a skeptic to
> prove a negative!

Here too, if you could say a little more. The point is not a question
of true and false, but of the consistency of related set of beliefs.
I maintain that comp ==> STRONG AI ==> MEC-BEH, and I don't
understand what is wrong with that.

Have a nice week-end,

Bruno


http://iridia.ulb.ac.be/~marchal/
Received on Fri Aug 26 2005 - 11:36:00 PDT

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:11 PST