[ExI] [Wittrs] Re: The System Level Issue

Gordon Swobe gts_2000 at yahoo.com
Mon Jul 5 23:37:05 UTC 2010


--- On Mon, 7/5/10, someone wrote:

> --- In Wittrs at yahoogroups.com,
> "someone else" <wittrsamr at ...> wrote:
> 
>> He is arguing that it doesn't make
>> sense to posit that algorithms (computations)
>> are of a natural kind enough to be the basis for
>> _explaining_ how consciousness may come to be.
> 
> I read this and the following questions run around in
> mind.

Without answering your questions in detail (I'll leave that to the person to whom you addressed) consider that computations have descriptive but not explanatory powers.

I might for example write a computer simulation that describes and illustrates in perfect detail a person P driving a nail of specification S with force F with a hammer of mass M into wood with density D. If my simulation takes all the relevant parameters then we will know with certainty the depth to which the person will drive the nail into the wood with his hammer. But does my simulation actually constitute a person driving a nail into wood? Or cause such a thing to happen? Of course not. 

Short of believing in some silly religion in which we all live as avatars in some glorified computer game, we must assume that the world exists before we make computations about it. It does not follow from the fact that we can compute process X that process X = a computation.

Does my digital simulation even _explain_ the process of man-hammer-nail-wood? No, it does no more than describe the process in terms of tentative and fallible hypotheses about the biology and physics of people driving nails into wood.

In other words, digital simulations or emulations of real process do not actually perform the processes they simulate. They serve only as models of or theories about those real processes. 

I think we can and should create models and theories about real things and processes, including real brain processes. I think we should not however conflate our models and theories with the real processes they're about. A model of a thing is not the thing modeled, even when the model is digital and the thing is a brain.

cc: ExI (relevant to the mind-body dualism thread)

-gts 








      



More information about the extropy-chat mailing list