Re: Challenge to the behaviourists, #1

From: AlphaOmega2004 (OmegaZero2003_at_yahoo.com)
Date: 09/15/04 

Date: Wed, 15 Sep 2004 13:48:19 -0700

"Sergio Navega" <snavega@intelliwise.com> wrote in message
news:10kh2l0ju38gu7d@news20.forteinc.com...
> "Wolf Kirchmeir" <wwolfkir@sympatico.ca> escreveu na mensagem
> news:YbX1d.5384$lb5.705717@news20.bellglobal.com...
> > Sergio Navega wrote:
> >
> > > "Chris Hooley" <hooleys@highstream.net> escreveu na mensagem
> > > news:acdeb196.0409141731.5faeeac4@posting.google.com...
> > >
> > [...]
> > >>Intelligence is internal. Behavior is external.
> > >
> > >
> > > This is by far the most important point in all this discussion.
> > > Definitions of intelligence using external (behavioral) aspects
> > > cannot distinguish between two organisms able to solve one
> > > "problem". One of them may be extremely intelligent, while the
> > > other may just "know" the solution. Possession of knowledge (or
> > > in other words, the ability to perform some action) is *not* an
> > > unequivocal demonstration of intelligence.
> > >
> > > Sergio Navega.
> >
> > If you can't tell the difference between "solving a problem" and
> > "knowing a solution", then they are the same, in which case no matter
> > what you call the "performed action", it's evidence of intelligence (as
> > defined by you.)
>
> That's the point, it's not. In my way of "defining" intelligence
> (at least for AI's purposes) one cannot use as a criterion the
> ability to solve a problem (or a specific performance, such as
> playing the violin). Using solely this criterion would include
> a cake whisk into the set of "intelligent" stuff.
>
> > Therefore the claim that what you have observed doesn't
> > prove intelligence proves merely that you have (let's be gentle here) a
> > vague notion of "intelligence".
>
> Thanks for the gentleness ;-)
> It is indeed vague, and I believe that definitions of intelligence
> considering just behaviors to be even more vague. Let's try to flesh
> that out. Here's the thing: I don't think that a system such as Doug
> Lenat's CYC is intelligent. I don't think that an Eliza-like
> chatterbot is intelligent either.

Since there are different types of intelligence, different attributes of
intelligence and most types are valued as within a range of values, (making
intelligence multifaceted), one can assign some tuple of values to such
devices/systems such as Cyc and Eliza without compromising what it means to
be intelligent.

>However, both of these systems may
> fool us (or some people) for a while. Both of these systems may
> present *external behavior* (in this case, verbal behavior) that
> leads one to (falsely) conclude that they are intelligent. I want
> to search for a definition that rule this out.
>
> For that reason, I want to work with the idea that intelligence
> is something that can only be determined once one have more
> information (even indirect) about what's happening *inside* the
> organism. I know that to say this to a behaviorist is to ask
> for rants. However, I don't see a way out.
>
> From the standpoint of someone that wants to *build* such things,
> it is *not possible* to avoid talking about the innards of such
> devices. AI is also an engineering effort. If this is inevitable,
> then it makes sense to use definitions that are consistent with
> this stance. My definition relates intelligence to the ability of
> an organism of generating knowledge. Obviously, now one is left
> with the task of defining what is knowledge, which seems to be
> a bit easier.
>
> Sergio Navega.
>
>

Relevant Pages

  • Re: Challenge to the behaviourists, #1
    ... > ability to solve a problem (or a specific performance, ... Using solely this criterion would include ... deciding this rather than "inside" means the organism is intelligent. ... "Generating knowledge" is neither consistent nor inconsistent with talk ...
    (sci.cognitive)
  • Re: Ability Goat, Absolute Ability Goat and Tennis Bolt
    ... Ability Goat and its ultimate version, Absolute Ability Goat ... postulation because I am presenting an objective criterion. ... THE FIRST SET results/dominance from semi and especially final of the ...
    (rec.sport.tennis)
  • Re: What is this thing called SELF?
    ... It is the ability in the entire vertebrate lineage to define ... the biochemical epitopes within each organism via the ... organism throughout it's life. ...
    (talk.origins)
  • Re: So I stumped you? Classes available to the ClassLoader...
    ... My take is that mutation is an artifact of the environment, ... Therefore the agent that mutates each organism will ... the ability to mutate itself, because that is an ability it never ...
    (comp.lang.java.programmer)