[Wittrs] Re: What the Man in the Room Knows (and when does he know it?)

  • From: "iro3isdx" <xznwrjnk-evca@xxxxxxxxx>
  • To: wittrsamr@xxxxxxxxxxxxx
  • Date: Thu, 25 Mar 2010 00:49:42 -0000

--- In Wittrs@xxxxxxxxxxxxxxx, Gordon Swobe <wittrsamr@...> wrote:


> Perhaps you don't understand that if computationalism = false then
> strong AI = false.

Whatever that means.

I worded my comment in terms of computationalism, because I don't  rule
out the possibility that people might someday try to construct  an
artificially intelligent system in some other way (i.e. not based  on
computation).  I prefer to wait until I see such alternative  methods
before I make any judgement on whether they are likely  to succeed.


> Searle only gives weak AI the benefit of the doubt when ...

And that's where Searle's thinking goes wrong.  He should have had  the
courage to disagree with the possibility of weak AI, instead  of
inventing his own mystical version (i.e. his "strong AI").

Searle elsewhere (I think the Scientific American article) says  that
intentionality is due to the causal powers of the brain.  The Systems
Reply is really just saying the same thing.  That is,  the only way to
get the behavior right is to get the causal powers  right, and therefore
if the system achieves weak AI (i.e. gets the  behavior right), it will
also have achieved intentionality by virtue  of having those causal
powers.

Searle's "response" to the Systems reply does not even address that
essential point.


> And by the way thought experiments can contain anything logically
> conceivable.

And they can produce bullshit conclusions.


> We can conceive of a Searle with a brain large enough and powerful
> enough to internalize and operate the program and pass the TT.

I can conceive of Chalmers zombies.  I can conceive of immaterial
substances.  These sorts of arguments are worthless.

The value of thought experiments is in helping people construct
suitable concepts to allow for discussion.  But such thought
experiments by themselves cannot constitute proofs.

Regards,
Neil

=========================================
Need Something? Check here: http://ludwig.squarespace.com/wittrslinks/

Other related posts: