On Feb 15, 1:22 pm, John Clark <johnkcl...@gmail.com> wrote:
> On Mon, Feb 13, 2012  Craig Weinberg <whatsons...@gmail.com> wrote:
>
> >> TO HELL WITH ELIZA!!!! That prehistoric program is NOT intelligent!
>
> > > What makes you sure it isn't intelligent but that other programs are?
>
> How the hell do you think?! ELIZA doesn't act intelligently but other
> programs do. Nobody in their right mind would use ELIZA to help with
> writing a scientific paper and doing serious research, but you might use
> Watson or Siri.

Obviously Watson or Siri give you access to intelligence, but so does
a book. Would you say that an almanac is more intelligent than a book
of poems? Does the IQ of a book change when you turn it upside down?
I'm trying to point out that what you associate with intelligence
figuratively does not correspond to literal capacity for intelligent
reasoning.

>
> > 20mb of conversational Chinese might be enough to pass a Turing Test for
> > a moderate amount of time.
>
> Maybe, if a chimpanzee were performing the test.

Yes.

>
> > It's completely subjective.
>
> Yes the Turing Test is subjective and it's flawed. Failing the Turing Test
> proves nothing definitive, the subject may be smart as hell but simply not
> want to answer your questions and prefer to remain silent. And
> unsophisticated people might even be impressed by a program as brain dead
> dumb as ELIZA. And people can fool us too, I think we've all met people who
> for the first 10 minutes seem smart as hell but after 30 minutes you
> realize they are pretentious dullards.

That's my point. But eventually we do realize they are dullards - or
machines.

>So with all these flaws why do we
> even bother with the Turing Test? Because despite its flaws it's the ONLY
> tool we have, its the only way of determining intelligence from stupidity,
> but if we are not very smart ourselves we will make lots of errors in
> administering the test.

You might consider that we don't need a test. That intelligence is
fundamentally different than muscle strength or height.

>
> > If you haven't read it already, this link from Stephen may do a better
> > job than I have of explaining my position:
>
> >http://newempiricism.blogspot.com/2009/02/symbol-grounding-problem.html
>
> And that fails the Turing Test because the author clearly thought that
> Searle was a pretty smart man.

He doesn't have to be smart to be right about the Chinese room. Even
if possibly for the wrong reason.

>
> >> You ask the room to produce a quantum theory of gravity and it does so,
> >> you ask it to output a new poem that a considerable fraction of the human
> >> race would consider to be very beautiful and it does so, you ask it to
> >> output a original fantasy children's novel that will be more popular than
> >> Harry Potter and it does so.
>
> > No. The thought experiment is not about simulating omniscience. If you
> > ask the room to produce anything outside of casual conversation, it would
> > politely decline.
>
> If that's all it could do, if it just produce streams of ELIZA style
> evasive blather then it has not demonstrated any intelligence  so I would
> have no reason to think it was intelligent so I would not think its
> conscious so WHAT'S THE POINT OF THE THOUGHT EXPERIMENT?

It would demonstrate X intelligence for t duration to Z audience.
Which is all any intelligence could hope to accomplish.

>
> > First you say 'let's say that the impossible Chinese Room was possible'.
> > Then you say 'it still doesn't work because the Chinese Room isn't
> > possible'.
>
> What I said was that real computers don't work anything like the Chinese
> Room, they don't have a copy of Shakespeare's Hamlet in which the letters
> "t" and "s" are reversed (so be or nos so be shas it she quetsion) resting
> in its memory just in case somebody requested such a thing, but if it had a
> copy of the play as Shakespeare (or Thaketpeare) wrote it simple ways could
> be found to produce it.

That's helpful but it is still the programmer's intelligence that is
reflected in the program, not the computer's. Which is the whole
point.

>
> > > The Chinese Room is just [...]
>
> There you do again with the "is just".
>
> > 'Where were you on the night of October 15, 2011'?
>
> Well, your honor my brain was inside the head which was on top of the body
> knocking over that liquor store, my mind was in a lingerie model's bedroom,
> and then on the moons of Jupiter. My sense organs are always very close to
> my brain but that is just a Evolutionary accident resulting from the fact
> that nerve impulses travel much much slower than light and if they were far
> from my brain the signal delay would have severely reduced the chances of
> my ancestors surviving long enough to reproduce.

Ah, then you shouldn't mind if we put your body in prison.

>
> > > There is a difference between organized matter and matter that wants to
> > organize.
>
> Carbon atoms want to organize into amino acids and amino acids want to
> organize into proteins and proteins want to organize into cells and cells
> want to organize into brains, but silicon atoms have no ambition and don't
> want to organize into anything??

They organize into silicon crystals but that's about it apparently.
Maybe they only need a helping hand, but so far there is no sign of
any grander ambition at all...unless they are really fukkin crafty and
are using the monkeys for muscle.

> Do you really think that line of thought
> will lead to anything productive?

Do you think that productivity for the sake of productivity will lead
to anything meaningful?

>
> > Why wouldn't he [Einstein] be aware of his own intelligence?
>
> You tell me, you're the one who believes that intelligent things like smart
> computers are unaware of their own intelligence.

If computers were intelligent, then yes, they would be aware of their
own intelligence to some extent, but since they aren't literally
intelligent, I would not expect that they are aware of anything beyond
voltage and current.

>
> > > We don't have to imagine solipsism just because subjectivity isn't
> > empirical.
>
> But that only works for you, the existence of other minds can only be
> inferred through behavior.

That's an assumption. You rule out all other epistemology arbitrarily.
I don't need to understand meteorology to know it's about to rain.

>
> > You admit then that you are not interested in defining it [intelligence]
> > as it actually is, but only what is convenient to investigate.
>
> Convenient? If intelligence does not mean doing intelligent things then I
> don't see why anyone would be interested in it and don't even see the need
> for the word.

Intelligence doesn't have to 'do' anything intelligent, it can just
make sense of things for the benefit of the subject.

>
> > You can't water corn with sulfuric acid
>
> You can if you change the organization of the acid a little. Sulfuric acid
> is H2SO4, remove the sulfur and 3 oxygen atoms and the result is H2O, and
> you can water corn with water just fine. In a similar way the only
> difference between a cadaver and a healthy person is the way the atoms are
> organized.

By that reasoning, all matter is the same thing. Gold, water, food,
poison...all meaningless configurations of quanta. Which is true in
some sense, but it is the opposite of cosmos, consciousness and
realism.

>
> > Organs produce fluids and heat, not consciousness.
>
> But something must and if its not matter then you must believe in the soul.
> I don't.

That's a false dichotomy. We don't have to be backed into a corner
just because we are only familiar with two approaches - we can try a
new approach which does not require belief at all, we can let reality
be what it actually seems to be. No pseudosubstantial 'soul' is
required, only ordinary identities.

>
> > Shooting someone in the head is not a software upgrade.
>
> That depends entirely on the person, I believe a recent bullet in the brain
> rather dramatically improved the performance of Osama bin Laden.

Figuratively valid maybe, but not literally true.

Craig

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to 
everything-list+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/everything-list?hl=en.

Reply via email to