Familiarity breeds contempt.

Once you understand each and every
instance of "intelligence", you may say
"it's not a big deal". It will take you
a longer time, and still more insight,
to appreciate intelligence again.

I'll write (or I will not) a series of posts
in the old good Chinese style of
presentig ideas via concrete examples.
I will also remember about yin & yang,
and I will ad a tinge of Greek too.

***

Consider the following problem:

there is a list of 100 integers

1231 1237 1249 1259 1277 1279 1283
1289 1291 1297
1301 1303 1307 1319 1321 1327 1361
1367 1373 1381
1399 1409 1423 1427 1429 1433 1439
1447 1451 1453
1459 1471 1481 1483 1487 1489 1493
1499 1511 1523
1531 1543 1549 1553 1559 1567 1571
1579 1583 1597
1601 1607 1609 1613 1619 1621 1627
1637 1657 1663
1667 1669 1693 1697 1699 1709 1721
1723 1733 1741
1747 1753 1759 1777 1783 1787 1789
1801 1811 1823
1831 1847 1861 1867 1871 1873 1877
1879 1889 1901
1907 1913 1931 1933 1949 1951 1973
1979 1987 1993

You need to provide the following final result (the total
accumulated sum):

(i) multiply the consecutive member from the list by 7;
(ii) multiply the result by 11;
(iii) multiply the rsult by 13;
(iv) add the result to the accumulated sum
(thus forming the new consecutive accumulated sum).

After you are finally done with the steps (i)-(iv)
for last item 1993, you're done. (At the start, the
initial accumulated sum is, by definition, zero).

***

So, that's your problem. It's like a problem for
a gofer: take this package and go West one mile
to the fifth street intersection; then turn North, go
a quarter of a mile to the 2nd intersection; then
turn East, go for a mile until you see the familiar to
you building with the "ABC headquarters" sign.
Deliver the package there.

***

Lets go back to the arithmetic problem.
An arithmetic gofer will do three multiplications
(i) (ii) (iii) per item, i.e. 300 multiplications,
and 99 additions.

An intelligent accountant may multiply

7 * 11 * 13 = 1001

first, Then s/he will multiply each item
only once, by 1001, and finally will
perform 99 additions.

***

A more arithmetically advanced accountant
will first of all add the numbers in the list
right away, which is 99 additions on smaller
nambers than in the earlier cases. And only then
s/he will multiply the resulting sum by 1001.

Thus 99 additions and 3 multiplications
suffice to solve the problem (instead of
300 multiplications, and 99 additions
on larger numbers).

***

Conclusions:

=====

One should avoid the most common weakness
of people working with in applied fields
(engineers, applied researchers and others):

do not (automatically) follow the formulation
(definition) of the problem. Reformulate it.

For instance, in the second problem above,
there is a possibility, that our gofer instread
of going 2.25 miles may walk straight up
North just a quarter of a mile.

The formulation of a problem often suggests
a solution. Often such a suggested solution
is lousy though or even impossible to follow
(while the people working on the solution
struggle and struggle in vain). The formulation
of the problem just identifies the problem, but
it does not have to be related to any reasonable
solution. (In this sense the formulation
may be misleading).

=====

The arithmetic presentation of the arithmetic
problem above looks as follows:

7*11*13*1231 + 7*11*13*1237 + ... + 7*11*13*1993

where we should actually write in all 100 summands,
not just three.

The advanced accountant's solution admits
a somewhat shorter expression:

1001 * (1231+1237+ ... +1993)

A common computer science wisdom tells
us that there is, as a rule, a trade off between
the time a solution takes (the number of operations
or computer instructions) and the memory needed
for a solution. (Chess players know it too, hence they
try to remember a maximal number of patterns).

The rule holds for OPTIMAL solutions, i.e. for
roughly equally intelligent solutions. As a rule, though,
a better programmer will write a program, which
is shorter, which uses less memory for the data,
and which is faster on the top of it (not to mention
better readability and documentation :-). Thus in the
above arithmetical problem, the better (more intelligent)
solution introduces a shorter arithmetical formula, and it is
faster too.

Remember:

understanding is synonymous with data compression

(relax guys--this is just my old aphorism,
not a strict scientific statement :-)

=====

One could say that we are dealing here simply
with arithmetic associativity and distributivity
laws of natural numbers, and not with intelligence.
As I said -- familiarity breeds contempt.

The actual problem could deal not with the numbers
but with money or other items. One had to recognize
certain laws, and that's an achivement. And anyway,
while most of the Silicon Valley programmers would
reformulate the problem in stride, majority of the
programers in other places and/or working for the
governmentor for the govewrnment contractors,
would still perform 300 muliplies, along
the initial formulation of the problem.

=====

We saw a glimps of precomputation at work
(7*11*13 = 1001). That's an intelligent thing
to do! This in particular is and should be done
in the case of the chess engines, where precomputation
(e.g. checking all posible openings in advance
and storing them in the internal book; and similar
idea applies to the endings too; etc) plays an
essential role. it is an intelligent thing to do, it
also makes the program much more intelligent
(great scientists tend to have a huge precomputed
and even raw data base in their head).

=====

The arithmetic distributivity law used above
also has a symbolic (metaphoric) meaning,
wildly applied. Just think about it.

****

If nobody cares to provide a feedback, I will stop
bothering you with this sequence. If I get mostly
negative feedback (the anonymous cowards don't
count) telling me that what I am writing is obvious
(true :-) and false (not true) and useless (well, it
depends) then I'll stop.

And if I get rather positive feedback then...
then I don't know :-)

Regards,

Wlod

PS. Intelligence is related to life and even to
reproduction and survival... All this is a huge
theme. As I said at the start, I am not attempting
to catch and cover the whole theme. Nobody can,
and it's even more true in my case. On the other
hand we live in the era when we already have the
answers/definitions about what is life etc. They
are not perfectly mature yet but we already have
the basis and the outline. The key words are
the *universal* Alan Turing machines, the John
von Neumann automata, DNA+RNA, etc.

===

(-: Since intelligence is synonymous with life,
the intelligent programs or robots should be
considered as living creatures. Destroying them
should be punished as any other violent crime :-)

wh

Anders Thulin is his mean himself:

wrote:

Familiarity breeds contempt.

Failure to follow newsgroup conventions also does so.

This is rec.games.chess.* -- if you want to hold forth on
topics without chess contents, look for a more appropriate
forum.

Skip it. (I really meant this thread
for the **intelligent** participants,
who understand the relation of this
thread to the popular rgc* topic
of chess programming and similar).

===

With a contempt for the small mean guys,

Wlod

I remember back to a time (a long way back) when the strongest
commercial chess program was Mchess, or MchessProfessional.
The main thing I noticed which was different about this program
was the scoring of Bishops vs. Knights. While most other strong
programs rated them about equal, Mchess went crazy, putting the
Bishop almost midway between a Knight and a Rook! The Pro
version scaled this back to a more reasonable level, but still gave
Bishops a clear edge.

Then along came a program by Richard Land, called Genius.
His program fit on a single disk, including the opening book! It
was much, much smaller, and reached at least a ply deeper in
the same analysis time. As you say, the programmer somehow
shrank the program, while at the same time improving the results.


[Mean guys please stop reading here.]


As for intelligence, I clicked on a Web ad for Tickle.com, and
got roped into taking an IQ test. What annoyed me was not that
the test went on and on, while I was very tired from being up all
night, but that many -- if not most -- of the questions did not
really target intelligence, but rather, education. For example,
a test designed for six year olds might ask:

2 + 2 = ?

3 - 1 = ?

6 x 3 = ?

This is not "intelligence", this is simple arithmetic!

Many questions focused on other aspects of education, such
as vocabulary, logic, and of course, math. Now my point is
this: a natural born genius with a poor education cannot possibly
be expected to score well on such a test. Likewise, a dullard
who is properly "educated" in the answers to such questions
(who memorises the answers or whatever) will do reasonably
well. Tell a dimwit: "Rooks belong behind passed pawns" and
then give him a chess problem with a Rook and a passed pawn,
he will probably put the Rook behind it (even if it hangs the Rook).

I don't think one can equate education to intelligence, because
if they were the very same thing, we have wasted a whole extra
word for nothing!

In any case, the mean guy has a point: this is a newsgroup
for stupid chess players, not those precious few who have
real intelligence (they won't be found here, wasting invaluable
time!). Mean guy -- I thought I told you to stop reading? Go
away, you big old bully.

GetClub.com is in dire need of some intelligent help in
chess programming, as they appear to have taken the
approach of pasting "band-aids" after each successive loss
by their computer. I have no doubt alienated Sanny, the
head honcho at GetClub, by giving his program a series of
painful "spankings".


Intelligence is this, intelligence is that -- here is what one
Web site has to say as to what intelligence REALLY is:

a) the capacity to acquire and apply knowledge

b) the faculty of thought and reason

c) Chess Life *has* no standards [inside joke]

d) superior powers of mind



-- help bot

wrote:

(..)

(-: Since intelligence is synonymous with life,
the intelligent programs or robots should be
considered as living creatures. Destroying them
should be punished as any other violent crime :-)

Wlod,

You are beginning to sound like some sort of punishment-freak. Destroy
the 'daleks' even unto their monstrous organic living squidgy bit, but
don't go around the Louvre slashing & pouring acid on the 'mona' & other
great works - thks. Yours for ever & I did 'scan' the snipped stuff..

Wlodzimierz Holsztynski (Wlod) wrote:
Anders Thulin wrote:
This is rec.games.chess.* -- if you want to hold forth on topics
without chess contents, look for a more appropriate forum.

Skip it.

The whole point of having a hierarchy of newsgroups is to keep related
posts together. If everyone just takes the attitude of `I'll post
whatever I want, wherever I want and people who don't want to read it
can skip it', we may as well just have one group called `news'.

If you are going to post off-topic stuff which you think is,
nonetheless, going to be of interest, you could at least mark it as
off-topic so that people can easily filter it.


Dave.

--
David Richerby Poisonous Umbrella (TM): it's like
www.chiark.greenend.org.uk/~davidr/ an umbrella but it'll kill you in
seconds!