Finally Science Meets Computer Chess XI (Turing)
Phil Innes
pioneer of computers is at all relevant to computer chess. However, this
pioneer used chess as a means to develop his ideas, and in fact, wrote the
first chess program.
He also theorised on the basis for intelligent machines, and proposed
two ideas which carry his name, and to sufficiently differentiate them here
are summaries of both - one called the Turing Machine (or Engine) and the
second the basis for Artificial Intelligence, the Turing Test.
It is important to understand both ideas from Turing, because they are
still the de facto paradigm.
[base text, Paul Hoffman]
The Turing Test was Alan Mathison Turing's idea of how to determine if a
machine can think. The machine calls for a person and a randomly choden
person to be seperated from an interrogator, who asks them each an unlimited
number of questions through an intermediary. Turing thought that if the
interrogator failed to distinguish between the machine and the human being,
it meant that the machine was thinking. In other words, if the machine
passes for intelligent, is /is/ intelligent.
The Turing Machine idea emerged at Cambridge University where Turing was
a Fellow of King's College [1935] and where he was exposed to the
revolutionary development in physics that toppled traditional ideas of
causality and determinism. According to the Newtonian worldview, if
sufficient information is known about a physical system, its entire future
can be predicted.
vis Pierre-Simon Laplace [1795] "Given for one instant an intelligence
which could comprehend all the forces by which nature is animated and the
respective situations of the being who compose itan intelligence
sufficiently vast to submit these data to analysisit would embrace in the
same formula the movements of the greatest bodies and those of the lightest
atom; for it, nothing would be uncertain and the future, as the past, would
be present to its eyes."
The introduction of quantum mechanics, however, in the early part of the
century, put an end to the idea that the future is completely determined by
the present and the past. Cambridge Univ. in the 1930'swas the center of
philosophical havoc engendered by quantum mechanics, particularly by the
principle that the observer always influences the observed. Turing found
this idea unsettling and turned to a perceived absoluteness in mathematics,
and consequently invented his Turing Machine (which is a mathematical model,
operating with or without a machine) which when mechanised utilises a
finistic [finite set of known numbers] to generate an infinite range of
numbers.
BUT
In the mid 1970's serious problems emerged around both Turing's ideas.
Challenges emerged from John Searle, a philosopher at Berkeley, Ca, from
John Lucas, a philospopher at Oxford, Eng, and another Berkeley professor,
Hubert Dreyfus.
This trio, each in a different way, challenged the explicit behaviors of the
use of computers by scientists, and 'social scientists,' ie, use of the
Turing Engine
And they also challenged the implicit assumptions of the Turing test, and
brought it under scrutiny.
Two branches of exploration into thinking machines have been categorised as
weak- and strong- AI. Strong AI carried many analogous features to human
thinking, and was indeed, the brave new frontier of exploration.
John Lucas was able to make a richer comparison between people and the
engines; showing, as a prime exhibit, the way that people learn in tiers,
that they gain information then transform that same information into
knowledge, then proceeded to repeat the process with the same or new
information, again transforming that information into knowledge by
successfully abstracting it as before, or to an even greater level of
complexity or amplitude.
He further showed that unless this process of transforming data took place
in human beings, the addition of more information would have only a minimal
effect on subsequent comprehension. There were periodic cycles to learning!
When comparing this behavior in human learning to the computer, only the
most perfunctory levels of information acquisition were evident, and the
function which in human being we call assimilation, is entirely absent in
the machines. [ for anyone who would want to read further John Lucas
somewhat took exception to Godel's work of 1931]
Dreyfus has posited that this lack of digestion of data is a direct result
of the emulatory paradigm as in the Turing Test, which only intends to seem
as if intelligent, and that scientific efforts in studying a conjunction of
human learning, memory and logic and computer science have scarce yet begun.
Together, these three caused such a controversy in academic and research
communities that very few AI claims are now made that escape further tests
proposed and designed by them to reassert other than the surface gloss that
Turing proposed - indeed, to requalify the old chestnut - a test of
intelifence is to determine the similar from the same.
________
Please excuse this very long note. Anyone interested in these writers can
probably find on-line resources, and all are published, if only in academic
journals.
The real implications for computerchess are thereby different that the one
model that has been deployed since 1985. It is as though the subject has
been worked on one side only, ignoring the other for the dubious gain of
appearing to be superior than in fact can be established by undertaking the
same activity as human players.
Commercial pressures on chess to make winning products are understandable,
but is not a sufficient reason to justify the extent of the diversion of
real effort to the seeming effort.
Chess was the first test-bed for ideas in AI, in theory and practice. For
those with an interest in scientific method and logical process who can work
together, it is necessary to comprehend Turing, and to step beyond that
lonesome and tragically worldy-unwise pioneer, and together develop the full
possibilities of machine intelligence and chess.
The possibilities for chess remain as indicated by our Dutchman, de Groot.
The real study is man.
Cordially, Phil Innes
Paul Onstad
-Paul
Phil Innes
> From: Paul Onstad <pon...@visi.com>
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
> I have spent the morning reading John Searle. Nothing to report yet.
<laugh> good! may take a while to digest.
when i read Searle and Lucas it was re-assuring that such essential,
fundamental and critical intelligences still operated
i have often been struck, from a position of feeling demoralised by either
popular writing or some zeit which had no geist whatsoever, by a writer who
brilliantly penetrates the miasma
not really in contradiction of others, but, and much better, in celebration
of what he has found and now expresses of the essential thing
cordially, phil
Peter Berger
> It may not be immediately evident why some exercise invented by a
> pioneer of computers is at all relevant to computer chess. However, this
> pioneer used chess as a means to develop his ideas, and in fact, wrote the
> first chess program.
The first chess program was written by Konrad Zuse in Plankalkuel
between 1942-45. You can even play against it on the Web as it has
been ported to Java and is availlable as an applet.
http://www.zib.de/zuse/Inhalt/Programme/Plankalkuel/Chess/JavaApplet/chess.html
Regards,
pete
Phil Innes
> From: pete...@gmx.net (Peter Berger)
> Organization: http://groups.google.com/
> Newsgroups: rec.games.chess.computer
> Date: 15 Nov 2001 05:07:37 -0800
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
> Phil Innes <aong...@sover.net> wrote in message
> news:<B81352D1.11A9%aong...@sover.net>...
>> It may not be immediately evident why some exercise invented by a
>> pioneer of computers is at all relevant to computer chess. However, this
>> pioneer used chess as a means to develop his ideas, and in fact, wrote the
>> first chess program.
>
> The first chess program was written by Konrad Zuse in Plankalkuel
> between 1942-45. You can even play against it on the Web as it has
> been ported to Java and is availlable as an applet.
hi peter, Turing wrote his program in 1937, it was a paper product, and
would have worked on a computer if there had been any computers!
in the first game ever played the program's Queen went on a pawn hunt and
got itself trapped :(
cordially, phil
>
http://www.zib.de/zuse/Inhalt/Programme/Plankalkuel/Chess/JavaApplet/chess.h
tm> l
>
> Regards,
> pete
Paul Onstad
>
> > From: Paul Onstad <pon...@visi.com>
>
> > Subject: Re: Finally Science Meets Computer Chess XI (Turing)
> >
> > I have spent the morning reading John Searle. Nothing to report yet.
>
> <laugh> good! may take a while to digest.
>
> when i read Searle and Lucas it was re-assuring that such essential,
> fundamental and critical intelligences still operated
>
> i have often been struck, from a position of feeling demoralised by either
> popular writing or some zeit which had no geist whatsoever, by a writer who
> brilliantly penetrates the miasma
>
> not really in contradiction of others, but, and much better, in celebration
> of what he has found and now expresses of the essential thing
I was trying to get a synopsis on "original intentionality" but got lost in
the process. I have referenced Searle on other matters, however, such as his
"Chinese Room" (rgcc discussion last January, AI and Kasparov).
He appears often in criticism for the New York Review of Books and this 1982
essay (days of the PDP-11) is still current on matters of computing:
http://www.nybooks.com/articles/6628 (chess is mentioned).
The AI debate is just one part of a much larger battle taking place in
science in which the same names keep popping up again and again. If all is
algorithmic, then so too is culture. ...That's Daniel Dennett, called by
Stephen J. Gould the lap-dog of Richard Dawkins :) Dennett in turn insists
that both Gould and Searle are believers in "skyhooks"--not the grounded
scientific necessity of "cranes." There are also "greedy reductionists,"
"spandrels," "cultural zoos" and now, "memes" (which I spotted Eric using
the other day). Search on any of them and the debates can be found.
Paul Feyerabend: "Scientists are not content with running their own playpens
in accordance with what they regard as the rules of the scientific method,
they want to universalize those rules, they want them to become part of
society at large, and they use every means at their disposal -- argument,
propaganda, pressure tactics, intimidation, lobbying -- to achieve their
aims."
-Paul
Rolf Tueschen
This witty quoting reminds me of an old experience. Two weeks before his final
(oral) examination for his doctorate our teacher took some days of being
indisposed and he told us that he had to learn about Marxism to begin with
because that was the modern topic of the time. Now I shrigged, because I knew
that you couldn't "learn" Marxism in a couple of days. Impossible to do. It's a
whole Science. The books of Mearx and Engels alone were some fifty in a row.
But what does it teach us? Simple. That also in Science a tendency of
impostering is well known. What is to be seen from the outside, is not the same
what is happening inside. Inside it's a certain predefined ritual. What is
examined, that is the cleverness you had shown in adapting to the internal
rules of the game. If you know the opinions of your examinor, you can learn the
necessary expressions in less than a week. Smarter would be in one afternoon.
But what I wanted to tell you, that is the impossibility to present yourself
(as being an engineer) as a real scientist just by reporting the reading of
some science scientists as Feyerabend and then quoting the ironical parts of
such an anarchist, that he was, as state of the art of a _fair_ description of
scientists, which it is NOT. <smile>
Rolf
Paul Onstad
Well, I wish I could share your optimism, but, while it might not be the
nice scientists you describe, there are all the other hangers-on who present
a real danger. To paraphrase Dennett: "Damn right we'll teach your kids
science and if you start muddling their little pointy heads with your own
ideas we've some cages set aside for you."
-Paul
Aaron
> Phil Innes wrote:
>>
>> > From: Paul Onstad <pon...@visi.com>
>>
>> > Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>> >
>> > I have spent the morning reading John Searle. Nothing to report yet.
>>
>> <laugh> good! may take a while to digest.
>>
>> when i read Searle and Lucas it was re-assuring that such essential,
>> fundamental and critical intelligences still operated
>>
>> i have often been struck, from a position of feeling demoralised by
>> either popular writing or some zeit which had no geist whatsoever, by
>> a writer who brilliantly penetrates the miasma
>>
>> not really in contradiction of others, but, and much better, in
>> celebration of what he has found and now expresses of the essential
>> thing
>
> I was trying to get a synopsis on "original intentionality" but got
> lost in the process. I have referenced Searle on other matters,
> however, such as his "Chinese Room" (rgcc discussion last January, AI
> and Kasparov). He appears often in criticism for the New York Review of
> Books and this 1982 essay (days of the PDP-11) is still current on
> matters of computing: http://www.nybooks.com/articles/6628 (chess is
> mentioned).
>
> The AI debate is just one part of a much larger battle taking place in
> science in which the same names keep popping up again and again. If all
> is algorithmic, then so too is culture. ...That's Daniel Dennett,
> called by Stephen J. Gould the lap-dog of Richard Dawkins :)
That's after Dennett devoted one whole chapter to attacking Gould in
"Darwin's dangerous idea"..including one devasting quote from John Maynard
Smith..
Which BTW is one of the most brillant and difficult to digest books that I
have read..
> Dennett in
> turn insists that both Gould and Searle are believers in
> "skyhooks"--not the grounded scientific necessity of "cranes." There
> are also "greedy reductionists," "spandrels," "cultural zoos" and now,
> "memes" (which I spotted Eric using the other day). Search on any of
> them and the debates can be found.
Hmm so how does this match into the Computer Chess?
Are the "new paradigm" gang looking for skyhooks? Or are the traditional
computer chess school "greedy reductionist"? :)
Just a thought
Paul Onstad
That would depend upon where the porridge is "just right." If a little on
one side, then the greedy reductionists are those building chess programs
like Heathkit radios. If on the other, then Rolf is a surely dropping his
hook from....a long way up :)
-Paul
Phil Innes
> From: Paul Onstad <pon...@visi.com>
> Newsgroups: rec.games.chess.computer
> Date: Sat, 17 Nov 2001 10:10:01 -0600
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
>> Are the "new paradigm" gang looking for skyhooks? Or are the traditional
>> computer chess school "greedy reductionist"? :)
>
> That would depend upon where the porridge is "just right." If a little on
> one side, then the greedy reductionists are those building chess programs
> like Heathkit radios. If on the other, then Rolf is a surely dropping his
> hook from....a long way up :)
please be careful of the transposition of terms:-
up is not opposite from down in any quantum sense, [Paul!!]
up and down are polarities of one axis
in a quantum sense down =very small number of setA; up=very large number of
setA: 0/1 and 1/0 is an extreme and weird expression, rendered respectable
algebraically by A/1 and 1/A
when i tested what people understood by this expression 0/1*1/0 here there
were many wonderful answers! the point in a _quantum_ [field!!] sense is
that there are no absolute values which can be stated without reference to
another value.
so answers of €0 or €1 or €infinitiy are meaningless.
in a quantum sense there are very few 1s in 0, and very many 0s in 1,
so we have a very small number interrogating a very large number
there can be a result(s!) only when other criteria are applied, then the
other criteria influences the result(s) in the field provided by 0/1:1/0
*meta-note 7, see below
____
rolf can be coming from another axis, setB; from deep [integration of chess
knowledge] to surface [performance]
of course, where these axies intersect is the right-angled centre of maximum
stress, and if one is immersed in one polarity set, the encounter may seem
obscure or irrelevent
however, it is interesting that [practically without exception] all recent
breakthroughs in science and application technologies for the past 25 years
have emerged from interdisciplinary studies, typically <stress> involving
recogition of these dual axies
__
prime example is in AI research where researchers have difficulty organising
or even finding frames through which to view their results , and have turned
to biologists, with the result that modelling occurs around the mysterious
proportioning of structures observed in insects, ants and bees
[goodbye Turing, hello Konrad Lorenz!]
and the biologists cannot cope with such complex interactivity without the
aid of computer models to abstract mathematically (actually, better said is
'geometrically') from the apparent 'chaos' of living systems
[goodbye neo-Darwin, hello David Böhm!]
__
the dearth of anything significantly new in computerchess is veiled by the
motif of winning as apparently supported by ratings;
and not based within legal constructions of chess rules interacting with
the other polarity set represented by human beings;
as such an absolute absense* of concise benchmark testing of chess playing
engines against setB [setA:setB = field of possibilities] tells its own
story
cordially, phil
* :))
> -Paul
Paul Onstad
I was just saying Rolf was close to God but thanks for reminding me there
are other interpretations :) Whew!
-Paul
Rolf Tueschen
Paul,
first of all you show very good what I tried to explain. - - People could think
now that you must be a real philosopher of the Science of Science. Or of AI.
Second point!
You show very well that you still have not understood what the wrong-doing was
in the TCC, the whole history of commercial CC in the last two decades.
The wrong did not exist in the philosophy or such super elevated fields, no,
the sins are all situated in the m-e-t-h-o-d-o-l-o-g-y of computerchess as part
of computer sciences.
This has been discussed in most of the "Finally Science meets Computerchess'
parts.
The fact that you and Bob and most of the programmers are unable to understand
the critical points, is somehow ridiculous because most of the testing and
pompous "methods" are worth exactly ZERO=ZILCH=NADA because of the lack of
basis (no exact definition of the machine/prog itself etc.).
I feel a great regret, that all these talented people here will have to face
that their whole activity is worth nothing, because the academic people
commited high treason to their holy laws of Science, because they once had
started in a somewhat pure hobby to get into closer contact with the then super
computer monsters.
But then things have changed. At that moment the academics should have created
r-e-a-l basics for computerchess. IBM/DB team members had the final chance
prior to the match against Kasparov. But they didn't want to do that. They
perverted computerchess in a variation of sports like boxing. Period.
Paul, you are very wrong when you try to make jokes about the actual emptiness.
It's about the life and existence of dozens of people. Grow up! And help to
find better ways. But that won't happen if you begin to reflect philosophical
fairy tales. Begin with the fundamental methodological details.
Rolf
Phil Innes
> From: Paul Onstad <pon...@visi.com>
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
>> as such an absolute absense* of concise benchmark testing of chess playing
>> engines against setB [setA:setB = field of possibilities] tells its own
>> story
>>
>> cordially, phil
>>
>> * :))
>
> I was just saying Rolf was close to God but thanks for reminding me there
> are other interpretations :) Whew!
<laugh> God?!
the other polarity set [whateverB], if rejected, often seems like an
encounter with the devil, no!?
Tuefelization Syndromachy!
(see - i made a joke)
otherwise the field is not abandoned by science, tho' not much of it exists
in the newsgroup; if people were applying method, they would present their
findings methodologically‹ apso facto, this is mostly absent
before a method can be introduced there has needed to be a deconstruction to
identify, if possible, which functions are distinguishable from others, and
the elimination of such /a priori/ methods as are commonly applied‹
otherwise too much is 'determined' before any observation can take place (!)
doh! and experimenters place themselves in the position of others from the
late 18th century‹ to such an extent that the subject matter, chess,
immediately recedes if a comparison is made to it, and forthcoming instead
are rationals for process [a double-shift away from method and chess]
very early on you spotted something about Turing's Test‹ it's duration. it
is possible to say that for short periods of time, and limited exposure, the
chess machina can emulate a person's behavior by simply copying its result‹
however, it is evident when the book has stopped, and engine calculus takes
over, noticeable from both the speed and quality of resulting play, no?
eonically [for any full cycle] the machine does not pass the test. neither
without book=on [for example] does it emulate master chess, but instead the
sub-lunar world of tricks and traps, and 2-bishop-rules, double-pawn rules
and so on, <stress> of fixed & pre-conceived evaluation [2]
a comparable test is to play against an 1700 player, whose behavior in my
experience is the same!
a scientifically valid point that CAN be abstracted from current behavior of
chess engines is to note the extent that engine calculus can negotiate
dynamic-geometries [1] and also _specific_ areas where calculation is
inadequate [2] and why [3]
i know that you are interested in the subject from a science viewpoint, and
if even a not-very-special method can be introduced to the evaluation of the
robot then this immediately differentiates it from other forms of
presentation (!) and non-dependently
this last item is where the scientist takes a personal responsibility in
deploying impersonal study
cordially, phil
notes:
[1] this is a very interesting area of exploration, and a continuous factor
in general robotics
[2] returns to the general paradigm of fixed and pre-configured evaluation
in comparison to real-world conditions; a time/unknown-scape challenge
[3] here is the breakthrough zone, de Groot identified it, hardly anyone has
pursued it
Peter Berger
> > The first chess program was written by Konrad Zuse in Plankalkuel
> > between 1942-45. You can even play against it on the Web as it has
> > been ported to Java and is availlable as an applet.
>
> hi peter, Turing wrote his program in 1937, it was a paper product, and
> would have worked on a computer if there had been any computers!
>
> in the first game ever played the program's Queen went on a pawn hunt and
> got itself trapped :(
>
> cordially, phil
>
As far as I know Turing did this not in 1937 but in 1947 while Zuse
had already finished his work in 1945. It has been quoted in several
books that Turing wrote the first chess program ( on paper). It seems
this piece of information is wrong.
Regards,
pete
Phil Innes
>> hi peter, Turing wrote his program in 1937, it was a paper product, and
>> would have worked on a computer if there had been any computers!
>>
>> in the first game ever played the program's Queen went on a pawn hunt and
>> got itself trapped :(
>>
>> cordially, phil
>>
>
> As far as I know Turing did this not in 1937 but in 1947 while Zuse
> had already finished his work in 1945. It has been quoted in several
> books that Turing wrote the first chess program ( on paper). It seems
> this piece of information is wrong.
pete, you may be right, if so, i have faithfully transcribed a typo from a
reference book which also gave the player's name and the gamescore. i don't
have the book, but will look it up at next opportunity and repost
in the meantime - you have it!
cordially, phil
> Regards,
> pete
Paul Onstad
I think many of us, at least, got the flavor of what you are saying (and
have said before). The thing is (and I'm not even a chess engine programmer)
I don't believe the languages being used (like C) offer themselves to the
direction you have implied. ..Or is it just that most read Knuth and not de
Groot? (mentioned below) ..Don't know as I've never even heard of the
latter. Surely someone will come along but will they be interested in chess?
In the meantime it might be worth the doing to incorporate a little in
acknowledging a criticism. Schroder seemed a bit interested....and what's to
prevent your allies, Hammer and SuneF from programming in all the wisdom you
and Rolf possess? ;) I think the difficult part comes when we look at some
code which has attempted anything "human" at all. It's just a mass of
conditionals, inelegant, and fragile. It might be "neat" but it's clunky
underneath. Someone needs to develop a new language first. I don't believe
the ones tried so far are up to the task.
-Paul
Rolf Tueschen
>and what's to
>prevent your allies, Hammer and SuneF from programming in all the wisdom you
>and Rolf possess? ;) I think the difficult part comes when we look at some
>code which has attempted anything "human" at all. It's just a mass of
>conditionals, inelegant, and fragile.
>Someone needs to develop a new language first.
Thanks for the demonstration or the of your misunderstanding. It's breathtaking
to observe the output of a blickering specialist. Now it's about a totally new
language. But we were talking about the _lack_ of any methodology. And we must
not invent new methodologies! It's enough to respect the _basic_ rules of the
existing methodology. Paul, you're completely misleaden or - another
interpretation of course - you're caught in a new act of intentional confusing
the audience. I tend to believe that the latter is right because you seemed a
bit too smart until now. But it could be an optical delusion on the net.
Rolf
Paul Onstad
I'll turn your own argument on you (even though the closest I got to
engineering was geology). Us simple programmers don't know what the hell you
are talking about. Okay, we do (a little) and we could probably fool you
with some code (for a little while) but all we would be doing was showing
our sympathy. That, you already have in my case at least.
(This is the same answer I just gave to Phil but worded differently.)
> I feel a great regret, that all these talented people here will have to face
> that their whole activity is worth nothing, because the academic people
> commited high treason to their holy laws of Science, because they once had
> started in a somewhat pure hobby to get into closer contact with the then super
> computer monsters.
If it is high treason I'm glad the jails are not set up yet but, nah, it's
just a little foolishness at most--like 99% of our occupations. Soon there
will be 1000 different chess programs but maybe something interesting will
still turn up.
If I were younger I'd probably have a GameBoy but what's so different with
chess?
> But then things have changed. At that moment the academics should have created
> r-e-a-l basics for computerchess. IBM/DB team members had the final chance
> prior to the match against Kasparov. But they didn't want to do that. They
> perverted computerchess in a variation of sports like boxing. Period.
Okay, and the solution's simple, just get rid of all this man/machine
nonsense because it's not here yet. Reform the situation by realizing the
fact and taking the comps out of competition. That's where the harm is being
done. No one should be forced to play a computer.
Now, if a player insists, it's fine. If computer chess interests become
sponsors--maybe--but watch them carefully. That's a potential area of abuse.
> Paul, you are very wrong when you try to make jokes about the actual emptiness.
> It's about the life and existence of dozens of people. Grow up! And help to
> find better ways. But that won't happen if you begin to reflect philosophical
> fairy tales. Begin with the fundamental methodological details.
I'm afraid I can't stop with the jokes. Those little demons come out whether
I want them to or not ); );
-Paul
Phil Innes
> From: Paul Onstad <pon...@visi.com>
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
>> and experimenters place themselves in the position of others from the
>> late 18th century‹ to such an extent that the subject matter, chess,
>> immediately recedes if a comparison is made to it, and forthcoming instead
>> are rationals for process [a double-shift away from method and chess]
>
> I think many of us, at least, got the flavor of what you are saying (and
> have said before). The thing is (and I'm not even a chess engine programmer)
> I don't believe the languages being used (like C) offer themselves to the
> direction you have implied.
oh‹ any language is okay, i started with punched cards, streaming tape and a
room-sized modem. the 'program' was the operating system! and you got to
write that too! 'language' is only instruction at a certain level of
convenience to the programmer, when very fast activity is needed the
language is 'unfriendly' machine code, when high interactivity it is slower
and more user friendly and 'high level'
"Hal, make me a cup of joe, willya, and stop whistling that dumb tune"
"Certainly Paul, would you like your coffee near the hatch door?"
> ..Or is it just that most read Knuth and not de
> Groot? (mentioned below) ..Don't know as I've never even heard of the
> latter. Surely someone will come along but will they be interested in chess?
this has not been the point at all. the question has been, are programmers
interested in what they are emulating? that is, to what degree have they
paid attention to chess? if an interesting study is ignored, why do more
studies?
it may be particularly galling when results from other fields suggest that
linear extensions of progress have foundered everywhere, and if [as de
Groot] what chess programmers have done en masse [per SuneF's copy-code
posts] is follow the first third of deG's identification‹ a successive part
requires a greater attention to high level chess praxis
it would even be true to say that most chess programmers do not understand
high level chess - rather than that they leave it as an unimplemented method
<laugh>
programmers don't know what they don't know, and this state of affairs is
disguised by making non-resolved [new term! bob alarm!] additions to the
engine which apparently boost it's performance, and frigidly maintaining a
distance from the real game by not introducing the program into any
same:same comparison with real players
of course this is constructed on commerial hyperbole, and vicarious power
tripping through the mechanism of the program
> In the meantime it might be worth the doing to incorporate a little in
> acknowledging a criticism. Schroder seemed a bit interested....
sure, but not a contributor
>and what's to
> prevent your allies, Hammer and SuneF from programming in all the wisdom you
> and Rolf possess? ;) I think the difficult part comes when we look at some
> code which has attempted anything "human" at all.
i'm sorry‹ this is such a confusion of terms. is logic human? can humans
process in a linear way? can they sequence?
these are all functions of the computer, and in fact, not even in anyone's
imagination <lol> can the computer do something that is not human :)
so your approach offers a dichotomy which does not exist, nor can exist, and
in fact ‹as rolf takes you to task, your approach is /itself/ a problem
however, rather than suggest pre-emptively what might be done out of our
ignorance, is it interesting to see where we are? how will we do that but by
testing?
> It's just a mass of
> conditionals, inelegant, and fragile. It might be "neat" but it's clunky
> underneath. Someone needs to develop a new language first. I don't believe
> the ones tried so far are up to the task.
which task? to do what?
the idea is not to abandon what there is, but to examine it.
actually, the result may be shocking! various factors of program
understanding may be validated, others depredated. i doubt anyone here could
predict which or to what degree.
the discipline necessary would be to make some tests, then to evaluate them
chessically, after which ...
cordially, phil
> -Paul
>
>>
Paul Onstad
Humans invented logic but it's only a small part of what we are.
> these are all functions of the computer, and in fact, not even in anyone's
> imagination <lol> can the computer do something that is not human :)
>
> so your approach offers a dichotomy which does not exist, nor can exist, and
> in fact ‹as rolf takes you to task, your approach is /itself/ a problem
"The sense of" has many possible inversions. That's all you've really
demonstrated.
> however, rather than suggest pre-emptively what might be done out of our
> ignorance, is it interesting to see where we are? how will we do that but by
> testing?
Testing's fine but must follow coding. Coding in turn depends upon language.
If you can't explain it in language, you can't code it.
That's why I implied the necessity of some other computer language that is
not driven by algorithms but I know of no such language. In the meantime
there are objections but no solutions. Solutions are implied but they are
inexpressible. That's the point we're left at. I suppose anyone can try to
overcome this but I notice those who are most insistent aren't trying at
all. It's someone else's job.
> > It's just a mass of
> > conditionals, inelegant, and fragile. It might be "neat" but it's clunky
> > underneath. Someone needs to develop a new language first. I don't believe
> > the ones tried so far are up to the task.
>
> which task? to do what?
> the idea is not to abandon what there is, but to examine it.
>
> actually, the result may be shocking! various factors of program
> understanding may be validated, others depredated. i doubt anyone here could
> predict which or to what degree.
We could never have imagined much of what has taken place in computing over
a short history. Still, if software pioneers had been shown the screen of a
Mac back then they would probably have understood it. The concept could have
been expressed as well.
> the discipline necessary would be to make some tests, then to evaluate them
> chessically, after which ...
Yes? :)
-Paul
>
> cordially, phil
>
> > -Paul
> >
> >>
Phil Innes
> From: Paul Onstad <pon...@visi.com>
> Newsgroups: rec.games.chess.computer
> Date: Wed, 21 Nov 2001 10:43:24 -0600
> Subject: Re: Finally Science Meets Computer Chess XI (Turing)
>
>> i'm sorryã this is such a confusion of terms. is logic human? can humans
>> process in a linear way? can they sequence?
>
> Humans invented logic but it's only a small part of what we are.
yeah, and how we play chess
>> these are all functions of the computer, and in fact, not even in anyone's
>> imagination <lol> can the computer do something that is not human :)
>>
>> so your approach offers a dichotomy which does not exist, nor can exist, and
>> in fact ãas rolf takes you to task, your approach is /itself/ a problem
>
> "The sense of" has many possible inversions. That's all you've really
> demonstrated.
go ahead, develop the point
>> however, rather than suggest pre-emptively what might be done out of our
>> ignorance, is it interesting to see where we are? how will we do that but by
>> testing?
>
> Testing's fine but must follow coding. Coding in turn depends upon language.
> If you can't explain it in language, you can't code it.
What is tested is the program's chess ability. This can be explained in the
English language, but not necessarily understoodãthat has to do with chess
knowledge.
We could compare this with the factors which produce the move, the
programs's instruction, rendered to English if necessary, <right?>
> That's why I implied the necessity of some other computer language that is
> not driven by algorithms but I know of no such language. In the meantime
> there are objections but no solutions. Solutions are implied but they are
> inexpressible.
Not so. They are -->method langauge again--> 'implied' from existing study.
Example from de Groot: what attempt have we made towards Äb/abstract dynamic
patterning. We have not explored it we have not developed little syntax to
express it. This is quite different from supposing that it is inexpressible.
How do we know since it is not not attempted?
> That's the point we're left at. I suppose anyone can try to
> overcome this but I notice those who are most insistent aren't trying at
> all. It's someone else's job.
What do you mean? It is not delivered to the group as a happy solution. The
thinking of the old school is to project often crass futures.
Any real development is a group activity the need for it must be established
socially - those who have made singular attemtps are DERIDED by the old
world, because they have not won any glittering prizes.
CCC in particular is a forum devoted to optimising prospects in world Äa.
Bob Hyatt has personally gone on record at putting down in virulent and
lurid terms development initiatives which further what we propose. Bob is
apparently available every hour of the day to denouce such prospects,
courtesy U. Alabama.
>>> It's just a mass of
>>> conditionals, inelegant, and fragile. It might be "neat" but it's clunky
>>> underneath. Someone needs to develop a new language first. I don't believe
>>> the ones tried so far are up to the task.
>>
>> which task? to do what?
>> the idea is not to abandon what there is, but to examine it.
>>
>> actually, the result may be shocking! various factors of program
>> understanding may be validated, others depredated. i doubt anyone here could
>> predict which or to what degree.
>
> We could never have imagined much of what has taken place in computing over
> a short history. Still, if software pioneers had been shown the screen of a
> Mac back then they would probably have understood it. The concept could have
> been expressed as well.
>
>> the discipline necessary would be to make some tests, then to evaluate them
>> chessically, after which ...
>
> Yes? :)
yes - to do the work first, then speculate after. anything else is
insincere, no?
phil
> -Paul
>>
>> cordially, phil
>>
>>> -Paul
>>>
>>>>