[ExI] The symbol grounding problem in strong AI

[John Clark]

On Dec 26, 2009, Gordon Swobe wrote:

> the formal nature of programs becomes most obvious at the machine level, where the machine states represented by 1's and 0's have no semantic content either real or imagined.

A series of electrical impulses representing 1's and 0's came down a wire and into my computer, the machine then translated those 1's and 0's into the above words. You are insisting that those words have no meaning real or imagined. You may have a point.

> Understanding is associated with intelligent behavior, yes, but the two things do not equal one another.

You keep saying that over and over again but repetition doesn't make it true. If Darwin was right those two things MUST be true. You claim to be a rational man but ignore a mountain of evidence that shows your understanding of how the world works must be WRONG. If you are really rational then even if you don't understand how those two things could equal each other you must conclude that they DO equal each other. The evidence allows for no other conclusion.

> I don't claim computers have understanding. They act as if they have it (as in weak AI) but they do not actually have it (as in strong AI). 

And that's why I don't use the terms strong or weak AI, it's a distinction that quite literally cannot be made, not between computers and not between our fellow human beings. Perhaps some people have suffered a mutation that renders them no more conscious than a rock, however intelligent and sociable they appear to be; in fact that mutation would likely find its way into future populations as it would be no more detrimental than the mutation to lack eyes or pigmentation have for animals who have lived for thousands of generations in dark caves. The mutation would even be beneficial as the individual wouldn't be making something useless from nature's point of view and those resources could be used for something important that Evolution could actually see.

I believe my brain cells have better things to do than ponder this possibility very deeply because even if true it is the "weak" intelligent people's (or computer's) problem not mine. 

> how can computers that run formal programs have understanding? 

I take it that this is to be regarded as a Zen koan, like "What is the sound of one hand clapping?" or  
"Why does a man ask a question if he is determined to ignore the answer?" 

 John K Clark






-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20091226/d718a0ee/attachment.html>