Some critical reflections on computer chess

[Skybird]

Today I discovered that Chessbase has released the new Fritz, Fritz-13. It is so brandnew that it currently still is not mentioned on their English webpage, only on the German site (where they already have 30 minutes of video interviewing the developer on functionality).

That software has a new function, called live check, if i remember correctly. The engine will constantly send all anaylsis and calculation results to a live server wehre it all gets collected, from all users world-wide. That way, a giant databasde with analysis results get created, and the software on every user's system also will access this database and bring up the best results already stored there (best means: the deepest calculating analysis so far, for example 32 moves). Users are also enocuraged to invest separate time into this analysis thing, becasue they can "conquer" - says Chessbase - the top place for the deepest rooting analysis for every psoition stored on that server.

Mind you, we talk about ALL playing situations and positions that get created by any user during play.

Imagione a scientific project, and a world-wide community donating parts of it'S indiviual system'S CPU cycles into processing data from a scientiifc data research. They do something like this with SETI data, don't they, I seem to recall there was something similir rergarding genetics, but I am not sure. All nice and well, it makes sense, doesn'T it, do have such a project if the ammpount of raw data you have canot be processed by one university'S capacities alone.

But does it makie sense for a game? Is that the purpose why that game is being played?

PC software has revolutionised the way chess is played today, it has chnaged the professional understanding of top players, and also has a dramatic influence on amateur players. Chess has turned not only int he way pros prepare for tournaments, it has also chnaged by rewriting theory, it has cxhnaged by style, it has become much faster, direct, aggressive. Beside that, for many low-ranking amateurs, online chess got infested by cheating, by bad behaviour, by xyoung people focussing on tuning their PC's chess skills with newest sofware andf trying to play a "chess manager" themselves, leaving the duel to computers, and focussing on earning prestige for themselves becasue their system has "won", or they own the latest database, the latest engine version.

Now chessbase seems to have started to completely analyse the game to death. If they succeed in this, it will mean that in the future people playing against a computer will just play against a giant databse that already knows, there is no more room for surprises, for discoveries, for finding out something new, charting some terrain incognita. Not only does it invite people to shift even more focus away from actually mastering the game and towards competing in a competition over who gains how many name entries in toip analysis done by his system, it also creates a psychological change that I find difficult to underestimate: with every move you just step to a position that already has been analysed to death and that by knowledge already has decided your fate. It is as if you were playing a whole match by just repeating moves from a book on opening theory, following a variation that is 60 moves long, until the match is decided.

I feel extremely uncomfortable and disillusioned with this. Somehow, the magic gets stabbed from behind that way, it seems to me. He who makes a catalogue of all knowledge of the world, by that eliminates all room for fascination and surprises, mystrery and imagination. And ambition and motivation as well? I eould say that already is a problem with our modern culture alltogether, a problem that most people are not aware of, maybe do not even understand what you are talking about when you mention it.

There have beenm some strategy board games with a finite number of possible ways for matrches to go, that have been fuilly analysed. And as far as I know, all of them are dead in people's interest today.

I will not consider Fritz 13 for buying. Beside that, Fritz 11 and Hiarcs 10 already give me more headaches than I can bear, while offering spectacular 3D boards.