Website founded by
Milan Velimirović
in 2006

16:54 UTC
ISC 2021



Remember me

Forgot your
Click here!
to create your account if you don't already have one.

Rating lists


MatPlus.Net Forum General Can computers compose artistic problems? (1)
You can only view this page!
Page: [Previous] [Next] 1 2
(1) Posted by Darko Šaljić [Tuesday, Jun 14, 2016 09:56]

Can computers compose artistic problems? (1)

This is the title of today ChessBase article.
I call all you great minds to react and defend our art by giving comments on this site.
(Read Only)pid=14674
(2) Posted by shankar ram [Tuesday, Jun 14, 2016 15:51]

Yes, they can.
There are already several composers who have developed specialised programs which can search for sound problems, with specific combination of normal pieces, fairy pieces, fairy conditions and stipulations.
Some of these works could definitely be called artistic.
It currently takes a human hand to select the gems from the ordinary positions.
In the future, if not already present, we can expect some additional logic to be built into these programs so that they do some part of the selection themselves.
(Read Only)pid=14678
(3) Posted by Nikola Predrag [Tuesday, Jun 14, 2016 16:08]

No they can't, since they can't understand. But the human understanding can decrease towards zero to "match" the computers. It's a tragedy, but it happens regularly.
(Read Only)pid=14679
(4) Posted by Darko Šaljić [Tuesday, Jun 14, 2016 17:26]

Then, can I please order one #2 Lacny 2x3 set to play mechanism with white pawn key-move?
(Read Only)pid=14680
(5) Posted by shankar ram [Tuesday, Jun 14, 2016 17:56]

No, Darko!
They haven't got to that level yet..
But some simple themes are already present in the works that I have seen. Solving difficulty is also quite high in some of the problems.
Computers have entered into other "artistic" fields like music, so chess problems, with their inherent geometry, should be no exception.
(Read Only)pid=14681
(6) Posted by Nikola Predrag [Tuesday, Jun 14, 2016 18:42]

Sure, the poor bureaucratic perception may proclaim anything as "art". The best way to kill the art is drowning it in the "bureaucratic art".
(Read Only)pid=14682
(7) Posted by Darko Šaljić [Tuesday, Jun 14, 2016 18:54]

I think and hope it is imposible.
We will never have computer composing music like Bethowen, painting like Durer, writing like Borges or composing chess problems like Kovačević!
Their "artistic" products could only trigger perception of cold and emptyness without any emotional response.
And real art is all about emotions.
(Read Only)pid=14683
(8) Posted by Nikola Predrag [Tuesday, Jun 14, 2016 19:37]

Darko, provoking a "Pavlov's emotional response" is what kills the art. Just include puppies and babies and there's the "emotion" in a banal propaganda.
Art is the creation of a whole NEW world, an original idea realized in some medium. A mere imitation is just a patchwork of non-original features, the Kitsch that relies on the "Pavlov's emotional response".

Of course, the creation of "new world" is creating new emotions, and not just resembling some old experiences.
(Read Only)pid=14684
(9) Posted by shankar ram [Tuesday, Jun 14, 2016 20:12]

If you enjoyed and appreciated a problem, and later came to know it was "composed" by a computer, would your enjoyment and appreciation lessen?
(Read Only)pid=14685
(10) Posted by seetharaman kalyan [Tuesday, Jun 14, 2016 20:19]

Several decades ago when computers were starting to be used for solving, A.S.M.Dickens wrote about a complete set of helpmates with WK,WS, BK& BS found using computers. Computers speed and human programming skill has reached such a leveltoday, I am sure many composers are already using their computers to find many more gems. But as Shankar pointed out, it still needs a human mind to locate the gems from heaps of very ordinary.
(Read Only)pid=14686
(11) Posted by Nikola Predrag [Tuesday, Jun 14, 2016 21:07]

I've seen many boring non-original "problems", lacking any idea which at least looked as being made by a computer, despite bearing a human name.
But there are interesting problems, with the position&stipulation probably generated by computer, where the human author himself has recognized an idea. That idea was NOT created by the computer, but by the human being.
And a spectator actually re-creates the author's idea by comprehending the content.
(Read Only)pid=14687
(12) Posted by Alain Villeneuve [Wednesday, Jun 15, 2016 10:48]

Right, Nikola. And there are still more examples among studies !
(Read Only)pid=14688
(13) Posted by Darko Šaljić [Wednesday, Jun 15, 2016 12:10]; edited by Darko Šaljić [16-06-15]

Maybe I did not well expressed myself, but I thought that emotions and ideas together must be in the phase of creativity. Perception always depends on the level of knowledge, taste, artistic capacity and talent of the observer. But it is more your field, as I know you are not an artist only by chess bord.
But your term "emotional response" reminds me of the famous movie, not with a happy ending in this context, when Descartes falls in love with Rachel, despite her negative result on a given test with a similar name...
(Read Only)pid=14689
(14) Posted by Siegfried Hornecker [Wednesday, Jun 15, 2016 12:49]; edited by Siegfried Hornecker [16-06-15]

In all fairness, Chesthetica is developing. Some of the even earlier problems were horrible. Those are only bad. In some years, maybe something good comes out.
I had contact with Professor Iqbal, and I believe he can create it, it just requires a lot of work (but then computers also took a quarter of a century to beat OTB masters).
(Read Only)pid=14690
(15) Posted by Nikola Predrag [Wednesday, Jun 15, 2016 15:38]

Darko, you've introduced "emotional response" (post 7) and I agree in principle. The question is - what generates the emotions? - the original features of some piece of art or the features simply copied/plagiarized from earlier sources?

In an OTB game, computer has the unambiguous goals, to win or at least, not to loose. If there's no computable forced mate, a (dis)balance of the material could be computed and more subtle criteria could be programmed, but the ambiguity increases. Still, the final purpose is clear to a machine.

What a machine might recognize as the final purpose of a chess composition? Fulfilling the stipulation? That is achievable but the correct move-sequences which fulfill the stipulation are only a tool for presenting the idea. A human perception is required to recognize the abstract content which could make an idea.

A clever program might define some patterns recognizable to a computer which then might generate a position that matches a certain pattern. If the pattern presents an original idea, that would make a joint problem - imagined by a person and constructed by a computer.
In principle, that's already the case - computer helps the realization of a human idea.
(Read Only)pid=14692
(16) Posted by Hauke Reddmann [Wednesday, Jun 15, 2016 19:03]

Popcoooorn, Popcooooorn... :P

I can speak only for myself: If my pet octopus snatches
my chess board and by chance composes a #2, I won't
judge it other than by its artistic merits.
Same for a well-programmed computer. To quote a quote
that always works in this respect: Are you feeling
humiliated by the fact that a car is faster than you?

(Read Only)pid=14694
(17) Posted by Darko Šaljić [Wednesday, Jun 15, 2016 20:35]

It is not a matter of humiliation but a matter of purpose.
People can always compete with each other although there are machines that are superior to them.
But if our goal is to search for the ultimate truth in a game of chess or to create superior work of art (instead of competition), maybe we should all give up and let the machine to do whole work.
(Read Only)pid=14697
(18) Posted by Nikola Predrag [Wednesday, Jun 15, 2016 23:18]

Hauke, you must perceive the idea before judging how "artistically" it was realized in the chess medium. Is your octopus aware of the idea, or you should explain it to your pet?
(Read Only)pid=14698
(19) Posted by Darko Šaljić [Thursday, Jun 16, 2016 07:39]

Part 2:
(Read Only)pid=14701
(20) Posted by Evgeni Bourd [Thursday, Jun 16, 2016 10:05]

I don't understand why you consider computers to be unable to compose pretty problems.
Just like you teach somebody what is pretty a computer might be able to understand it by himself in the near future.
Lets ignore for the moment the "how" the computer composes, if you input him all the existing problems and their ranks he might be able to understand what we consider to be pretty ( long moves, inferences, flight giving and so on )
From that point if he follows the rules of our "pretty" you will probably not notice any difference, if anything the computer with its strength will always be better.
The attempts you see by people are rather poor to what you should expect a strong AI general theme learning computer, and that probably gives you a false impression to how general purpose computers might soon be similar to humans.
(Read Only)pid=14702

Page: [Previous] [Next] 1 2

MatPlus.Net Forum General Can computers compose artistic problems? (1)