🚀 go-pugleaf

In article <1992Sep18.172123.14367@mp.cs.niu.edu> rickert@mp.cs.niu.edu (Neil Rickert) writes:
>In article <7525@skye.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writes:
>
>>The net is wall-to-wall with people who insist on, and devote all
>>their intellectual abilities to supporting, the claim that some sort
>>of behavior (as in the Turing Test or Harnad's Total Turing Test) is
>>all that we should (or can!) count as evidence of consciousness.  Such
>>things as functional organization count for nothing, in their view.
>
>Either you are talking about a different net, or your interpretation of
>people's statements is very different from mine.  I see relatively few
>people who claim that TT is all we should count as evidence, or who
>claim that functional organization does not matter. 

Why is it then that when I suggest that functional organization 
(or, more generally, "how it works") should count, no one on the
anti-Searle side says they agree?  The response has been so
universally negative that I've more or less given up on trying
to make that point.  I used to post examples involving such
things as different kinds of Chess programs, to show roughly the
kind of thing I had in mind, and it never got me anywhere.

>             There are many more
>(and I count myself as one) who say that TT is all we can count, at
>least until we understand enough about the functional organization of
>the brain.  But most people in this category would say that functional
>organization is important, but since we don't know what functional
>organization is needed, for the present we will have to rely on TT as
>a guide, imperfect as that may be.

But we _don't_ have to rely on the TT.  There are no non-human
TT-passing entities that we have to deal with.

So why do people defend the TT?  There's no urgency about the issue.
It's not like _tomorrow_ we're going to have to decide whether to
give machines the vote.  It's not like _next week_ we're going to
have to deal with talking frogs.

Moreover, it should be straightforward to see that the TT is
fataly flawed.  A significant aspect of human consciousness is
that we condict an internal dialogue.  That is, we can think
thoughts to ourself in linguistic form.  I can think "I should
go home soon", for example.  Now, this is not necessarily
evident in behavior (mind reading from behavior is not possible,
so far as we know) much less in the sort of behavior that can
be seen via a teletype link.  So for this sort of question about
mental life the TT, and other behavioral tests, are almost
completely useless.

>I don't see anyone "devoting all their intellectual abilities" to
>supporting their claims.

Well, my exact words were hyperbole.  But we do see, for instance,
a number of people working to find flaws in Putnam's "rocks are
FSAs" arguments and then failing to do similar work against
claims about mappings between table lookup machines and the
brain.

And we do find people devoting a lot of time and intellictual ability
to the support of the TT in the attack on Searle.  Do you really think
this isn't true?

>>How they can be so sure that nothing we can find out about the brain
>>(for instance) will ever be relevant is beyond me.
>
>The brain is extremely complex.  It is quite likely that we will understand
>the brain only by building our own hardware and/or software and finding
>what kinds of functional organizations work, then using that as a guide
>to try to understand the brain.

I agree.

-- jd