From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wupost!bcm!aio!gothamcity.jsc.nasa.gov!dlyndes Thu Jan 16 17:22:00 EST 1992
Article 2743 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wupost!bcm!aio!gothamcity.jsc.nasa.gov!dlyndes
>From: dlyndes@gothamcity.jsc.nasa.gov (David Lyndes)
Newsgroups: comp.ai.philosophy
Subject: re: Cargo Cult Science
Message-ID: <1992Jan15.181608.20360@aio.jsc.nasa.gov>
Date: 15 Jan 1992 18:16:08 GMT
References: <92Jan15.081805est.14473@neat.cs.toronto.edu>
Sender: news@aio.jsc.nasa.gov (USENET News System)
Reply-To: dlyndes@deltahp.jsc.nasa.gov
Organization: Barrios Technology @ NASA/JSC; Houston
Lines: 81

In article <92Jan15.081805est.14473@neat.cs.toronto.edu>, mgreen@cs.toronto.edu (Marc Green) writes:
|> It's clear from the discussion that advocates of Strong-AI, and
|> computer scientists in general, don't have much understanding of
|> empirical science. The essense of science is refutability. For any
|> hypothesis to be taken seriously, it must be open to refutation.

Scientific theories are NOT (in general) refutable!  Because:

(1) In order to test a theory, you need to come up with some
      T  a theory
      I  initial conditions
      S  simplifying assumptions, then

(2) Calculate some expected observation, O

      T + I + S -> O
    
    The calculation itself is not necessarily a trivial matter.  It
    also may introduce factors which make "refutability" more
    problematic.  But we can let that one slide for now.

(3) Notice that if the expected observation O turns out to
    be false, we are NOT warrented in concluding that the theory
    T is false, only that the conjunction T + I + S is false.  By
    DeMorgan's rule, it only follows that

         T is false    OR    I is false    OR    S is false.

    We can also let it slide that taking the observation is also
    problematic, as any review of psychological literature will show.

A simple example; consider:

    T = Newton's theory of gravitation;  F = k m1 m2 / r^2

    I = Observe the planets and sun in sone initial configuration

    S = Assume that no other forces are acting on the planets and sun,
        there are no other planets, that they travel in a hard vacuum, etc.

    O = a calculated orbit for mercury.

The fact that the orbit calculated for mercury does not agree with the
observed orbit did not and does not refute Newton.  It is perfectly
reasonable to try to adjust I and S to get a better O, and this is
exactly what happened for a couple of hundred years.
 
|> If you point out that they have never come close to achieving any of
|> their goals of general intelligence, you simply get the someday
|> defense.

That is not all you get.  First: the strongest argument that AI is
possible is to produce AI.  Some opponents of AI demand (and rightly so)
of their AI counterparts that AI be demonstrated.

Second: In the mean time, we get arguments that AI is not possible.  It
is politically and intellectually prudent that supporters of AI defend
themselves.

Third: it is in the nature of a scientific enterprise to have some sort
of plan for going about achieving their goal.  This is called a "research
program."  There are lots of research programs in AI and there is a
shortage of funds.  So the researchers are prudent to argue that their
program has a reasonable chance of success.

|> This mind set results in stupid arguments like the one over Searle and
|> his simple-minded Chinese room.

Searle's argument is neither simple-minded nor stupid.  Inconclusive probably.
Wrong possibly.  But sophisticated, to the point and worth while in
context described above.

-- 

	 +-------------------------------------+
	 | David K. Lyndes                     |
	 | Barrios Technology                  |
	 | email: dlyndes@deltahp.jsc.nasa.gov | 
	 |       "It's my opinion and          |
	 |        you can't have it!"          |
	 +-------------------------------------+


