[ExI] Searle and AI

Gordon Swobe gts_2000 at yahoo.com
Tue Dec 29 01:35:17 UTC 2009


--- On Mon, 12/28/09, John Clark <jonkc at bellsouth.net> wrote:

>> Even if Zeus handed us a concrete example of an artificially
>> constructed machine with strong AI, we could not abstract
>> from careful study of that machine a formal program to run
>> on a software/hardware system that would enable that s/h
>> system to also have strong AI. We would need instead to
>> recreate that machine.

> Well I agree that if you duplicated a conscious machine the copy would 
> be conscious. 

Glad we can agree on something.

> Of course making a perfect copy of such a complex thing wouldn't be easy;
> to do so you would need a very long list of instructions
> specifying which of the 80 elements with a stable
> isotope you're dealing with, and information on
> where to obtain such such an atom, and the coordinates
> of where to move that atom to. There is a name for a list of
> instructions of that sort, it's called a, it's
> called a,..., oh damn, it's right on the tip of my
> tongue, give me a second, ... ah yes now I have it, it's
> called a PROGRAM.

Sure, we would most likely need a program like yours to guide the assembly. We might even build robots that run your program and put them to work on assembly lines for the purpose of building strong AI machines. But the program-driven robots would not have strong AI, at least not by virtue of your program.

However, in phase II of the project, the strong AI machines might take the place of the robots. (Unless they're lazy like me and would rather let the robots do it.)

-gts



      



More information about the extropy-chat mailing list