Re: Cool visual illusion


"feedbackdroids" <feedbackdroids@xxxxxxxxx> wrote in message
news:1133922291.081188.84070@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>
> Allan C Cybulskie wrote:
> > Then why did you say -- and I quote: "you would see that Wolf did
> > > > > exactly what hard-core behaviorism would never do, and what
philosophy
> > > > > is incapable of doing. Which is to attempt to explain outward
> > behavior
> > > > > by "going inside" to look at internal mechanisms."
> >
> > If you think that philosophy is capable of doing it, which is what I
said,
> > then why did you say above that it couldn't?
> >
> > Whether philosophy DOES or not may be another matter, but Stich is
certainly
> > one who would contradict any notion that philosophers DON'T look at the
> > internal mechanisms, or advocate such. So you're just wrong no matter
how
> > you slice it.
>
>
> The implication was that Wolf was being a physiologist when we made his
> post, not acting as behaviorist or philosopher.

Hmmm. So Stich doesn't act as a philosopher when he philosophically posits
that folk psychological terminology should be replaced by references to
specific neural patterns? I don't think he'd accept that.

Why can you never just simply admit that you made an overstatement? Your
own words are quoted above to prove that you did just that.

>
> The DUH was genuine shock that someone would actually propose that ...
> all explanations of behaviour should start and end at neurons. A real
> shocker.

I think I'm reading sarcasm here, but it is indeed a shocker since Stich
wants to replace terms like "X wants a cookie" with "these neurons are
firing", which most people find less than useful. Personally, I want to
know if they want a cookie, not that things are happening in their brain.

>
>
> > > He actually "used"
> > > the systems approach in the middle of his post,
> >
> > But you don't understand what "the systems approach" actually is. Hint:
It
> > isn't simply staring at a low-level implementation and thinking that
you've
> > done something cool.
>
>
> Had you ever viewed the List that we've been talking about for days
> now, you would understand what's wrong your statement here. On all 3
> points.

Um, since I REPLIED to the list, and in fact even was the first to suggest
that behaviourism could come into play at 1 as well as at 6, this statement
proves nothing more than that you are not paying attention.

>
>
>
> > >
> > > Philosophy, AT BEST, is verly loosely involved in item #1. Behaviorism
> > > at items #1 and #2 [and maybe again later during test+revision].
> >
> > Actually, you commented that behaviourism only came in at 1 and at 6, so
now
> > you are changing your story.
>
>
> Wolf originally said #6 - test and proof part. I agree it might help
> out there ... **IF** we ever get that far. For the foreseeable future
> however, maybe #1 and #2. One thing you have to realize is that, in
> practice, the boundaries between adjacent levels is a little fuzzy. One
> of the things you learn by actually "doing" it.

Yes, which is why when I commented on philosophy's place I said that it was
a bit in 1 and more in 2 and 3. And since I've never changed my story, that
seems to indicate that I get it better than you.

> Besides, your steps 1 - 3 are user
> > requirements, system specifications, and general design specifications.
> > Philosophy is much more involved in steps 2 and 3 than it is in step 1,
and
> > behaviourism is much more involved in step 1 than it is in step 2.
> > Basically, not only do you not know what you're talking about, you seem
to
> > not know what even you THOUGHT you were talking about. A most
disturbing
> > development ...
> >
>
>
> Quote: "... Philosophy is much more involved in steps 2 and 3 than it
> is in step 1..."
>
> Wonderful. Your turn on the block now. Beh had its day.
>
> What is YOUR philosophical proposal regards #3 = general design
> specifications? If possible, poach into item #4 a bit, if you will.
> Clock is ticking.

I claim that general design specifications gives us the algorithm for
implementing intelligence in a general level. My current best theory is
that the tenets of folk psychology need to be developed to give us the
algorithm for intelligence, since it actually currently is our best
method -- controversially for the behaviourist -- for predicting the
intelligent behaviour of people. So we need to flesh out and understand and
repair the concepts of beliefs, desires, rationality, etc, etc. But that's
philosophical work, not neurological work.


>
>
>
> > > There
> > > is probably *already* enough about behavior known to keep neuroscience
> > > busy for centuries.
> >
> > And so then behaviourism should just stop because neuroscience doesn't
need
> > it? Will you get it through your head that different people have
different
> > research interests and that just because you don't find theirs
interesting,
> > it doesn't make it wrong?
>
>
> Upside down thinking, bunkie.
>
> Beh can do as it likes. However, the OBVIOUS point implied was that,
> due to the fact that brain physiology is probably 1000X [rough
> estimate] more difficult than behavior observation, neuroscience will
> always lag far behind. Eg, it takes, say, 1B neurons to perform a
> simple behavioral act, but the act is observed and recorded by a
> behaviorist in a split second. Deciphering the underlying mechansisms,
> well ... you do the math.

But does this mean that behaviourists should STOP doing behaviourism?
Because that's what you implied and what I responded to.

> >
> > That being said, staring at an implementation isn't going to get us any
> > closer to that algorithm.
> >
>
>
> What implementation? We ain't got no really good implementations.
> That's the problem.
> DUH.

So you're claiming that the brain is not an implementation of intelligence?
Odd statement considering your stance ...

> > >
> > >
> > > Don't you know any biology at all?
> >
> > I'm asking you: Where on your set of 8 things in the systems approach
does
> > neurology enter the picture? Can you not answer a simple question?
> >
>
>
> You figure it out. I'm not gonna give you all the answers.

Sorry, but I am under no obligation to figure out your own position. YOU
insisted that behaviourism and philosophy weren't in the "right" places on
the list, and that neurology was. It's up to you to say where that actually
is.

> > > You're no programmer, are you.
> >
> > I've been working for 9 years on real production systems. What's YOUR
> > background?
> >
>
> Ok. Give some percentages. Add up to 100% for the total project.

Attempt to dodge noted.

Fine. Real world example. We started working on a feature in March. We
started writing code in September. The code is mostly done in December. So
that's about a 60-40 split between everything before implementation and
implementation. I won't use your rankings below because I consider -- as I
have said many times -- getting the algorithm to be part of number 3, but
"detailed design" as you mention it below counts both that and getting
classes, which I consider the specific design, and also you left out the
user requirements in your list (for shame). So it seems reasonable to say
that user reqs + general design = implementation + testing in terms of TIME.
In terms of brain work, user reqs + general design > implementation +
testing (testing is a fairly brainless task, but don't tell our test team I
said that [grin]).

> > > What you say here is what they teach
> > > freshmen in Pascal-101. If you think top-down, the code literally
> > > writes itself. Cool.
> > >
> > > Real programmers know a lot more. The lion's share of work - and time
> > > spent - comes in implementation, maintenance, and revision. And
> > > revision often [or usually] goes back to the topmost levels.
> >
> > REAL programmers are smart enough to know that once they have the
design --
> > ESPECIALLY with OO -- the coding, while still work, has nothing to do
with
> > the structure of the system. REAL programmers also know when they have
to
> > slip back to doing design during the coding phase.
> >
>
> Yes. Item #7 = revision. An enpanded lecture would have described how
> revision normally involves all levels. Sometimes you just cannot meet
> the customer reqs, no matter what.

Yes, but as I already said I would consider revising the user requirements
time spent in the user requirement section. Revision, to me, is a fairly
pointless category except for planning.

>
>
>
> > If you only believe that you can implement intelligence on a brain, then
> > stop doing AI. Building a brain is not building AI.
> >
> > What is it about brains that are so special that only THEY can cause
> > intelligence?
>
>
> Being the only known existence proof. Got some others?

This does not prove that they are the only things that can be intelligent.
In fact, computer science has pretty much shown that that is unlikely to be
the case. We can sort numbers on sequential processors, multi-threaded
apps, and hypercubes.

>
>
>
> >
> > > Did this
> > > idea ever occur to you? DOH.
> >
> > Why should it? I'm advocating that intelligence is an algorithm and
> > CONSCIOUSNESS is the result of non-physical forces. Get the arguments
> > straight.
>
>
> Give us your proof that non-physical forces exist. Don't tell us to
> disprove their existence. The monkey is on "your" back, if you want to
> be taken seriously.

I have given my evidence for my hypothesis. If you want to say that it is
impossible, then you do indeed have to disprove its existence. If you
simply want to say "I don't believe that" then we should have no argument.

>
> BTW, today I happened to pick up Dyson's book Infinite in All
> Directions. Ref pg 96, ... "Philosophy is nothing but empty words if
> not capable of being tested by experiment".

I have that book, actually, but will just note here that you who so despise
philosophical made-up stories seem to accept them awfully easily when they
support your position.


.

Relevant Pages

  • Re: How much intelligence?
    ... I like Curt, but my opinion is that he needs to ... He is trying to do philosophy without ... what philosophical problems behaviourism creates. ... But even here he got the definition of truth upside down ...
    (comp.ai.philosophy)
  • Re: the liver and the brain
    ... philosophy, since you seem to do the same thing. ... behaviourism. ... > Michaels can't see the extent to which he expresses his ... Lester and Dan and Eray may or may not be dogmatists, ...
    (sci.cognitive)
  • Re: How much intelligence?
    ... If behaviourism is true, not only is it ... Since it was the philosophers who STARTED science (science started from ... It's philosophy who'll tell us what limits science has. ... if one knows the truth choice in philosophy is not a matter of truth. ...
    (comp.ai.philosophy)
  • Re: How much intelligence?
    ... If behaviourism is true, ... Since it was the philosophers who STARTED science (science started from ... It's philosophy who'll tell us what limits science has. ... if one knows the truth choice in philosophy is not a matter of truth. ...
    (comp.ai.philosophy)
  • Re: How much intelligence?
    ... If behaviourism is true, not only is it ... Since it was the philosophers who STARTED science (science started from ... It's philosophy who'll tell us what limits science has. ... if one knows the truth choice in philosophy is not a matter of truth. ...
    (comp.ai.philosophy)