It is currently Thu Mar 28, 2024 2:02 pm

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 6 posts ] 
Author Message
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #1 Posted: Thu Aug 22, 2019 12:00 pm 
Honinbo

Posts: 10905
Liked others: 3651
Was liked: 3374
Marcel Grünauer wrote:
I'm wondering about the value of certain types of Go books in the age of ubiquitous superhuman-level AI.

- pro game collections, commented or not
- books on the opening
- Joseki books

AI can't explain the theory behind moves, but most of the time those kinds of books don't either.


Oh, I think that opening books will make a comeback. :) Humans will come up with new ideas and explanations. Also, in go you are thrown on your own resources early in the game, and understanding opening ideas can be a big help.

BTW, my latest posts in This 'n' That ( https://lifein19x19.com/viewtopic.php?p=247892#p247892 and https://lifein19x19.com/viewtopic.php?f ... 42#p247942 ) may be seen as notes for a book on openings. OC, more research is needed, but I think it likely that usually it is important to occupy the last open corner. Moreso than an open corner when there are more of them. :) That tidbit, if it holds up, may not be worth very much, but maybe 10% winrate. :) Also, it seems clear that the side is not worth as much as we thought just a few years ago. That's important. :)

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #2 Posted: Thu Aug 22, 2019 2:30 pm 
Lives in gote

Posts: 532
Liked others: 1
Was liked: 136
Rank: 13k
Marcel Grünauer wrote:
I'm wondering about the value of certain types of Go books in the age of ubiquitous superhuman-level AI.
I'm tempted to sell most of these types of books.


Do you mean the physical books or all that ancient stuff printed in them? I do not think your opinion is rare these days. Who is buying books? Your effort to recoup some money may be misdirected. Find a go club that has a lending library or a city librarian who will promise to put them into circulation on the shelves. Put the name and website of your go club on the back cover so the next generation of go enthusiasts can contact a human being for a game.

Physical books were supposed to have died off decades ago. Hasn't happened. Go theories come and go, rise and fall in popularity. I have no idea what this AI thing is all about since the things play at a level I cannot comprehend. But are we certain these algorithms are — really, finally, absolutely — the last word in go theory?

I was watching an interview with a pro the other day on the YouTooby thingy. He said (paraphrasing loosely), “I have played the computers. I lost. I wanted to win, of course, but I want to know we both played from the heart. The games were inelegant and lacked soul. The computers are for others. I do not need to play the computers again.”

I’m 15-10k, I don’t play online, and I don’t play competitively. I play to simply enjoy the game.

_________________
David Bogie, Boise ID
I play go, I ride a recumbent, of course I use Macintosh.

Top
 Profile  
 
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #3 Posted: Thu Aug 22, 2019 10:39 pm 
Judan

Posts: 6087
Liked others: 0
Was liked: 786
Pro game collections: If uncommented, the greatest competition to printed books is databases. Pro games do not become invalid because of AI but AI games are also worth studying. If commented, the commentaries are worth reading especially since AI hardly comments at all. Of course, one must be aware of new AI play.

Opening: The best books I know are the sparsely commented dicitonary by Rin Kaiho whose game sequences are interesting but the comments hardly, and the IIRC uncommented dictionary by the Nihon Kiin. Reading them is like reading uncommented pro games. See above for AI. The books are still very useful but one must, OC, also have an open mind towards AI play.

Josekis: AI has not rendered joseki knowledge and understanding useless. One must know the josekis before one can decide which to avoid or reinterpret in light of AI play. AI must not be an excuse not to learn josekis. The most useful joseki theory is in my joseki books. Most of it is still valid with these two restrictions: AI has introduced / altered a few josekis, especially 3-3 under 4-4; in josekis with territory / influence balance, AI has shifted it slightly towards territory, especially when there is potential for much (more) safe territory than in the initial joseki position. Therefore, this aspect of potential territory enhancement must be taken into account for my territory / influence ratio applied to some such josekis. The aspect is an advantage to be compensated by a ratio favouring the opponent instead of, as before, just having a balanced ratio. Such compensation is not new to my theory but, for those josekis, I (like everybody else) underestimated the advantage generated by aforementioned aspect.

Top
 Profile  
 
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #4 Posted: Fri Aug 23, 2019 6:17 am 
Gosei

Posts: 1625
Liked others: 542
Was liked: 450
Rank: senior player
GD Posts: 1000
I'm not getting rid of my books. In addition to what Robert mentioned, I see them as records of history of the game. As someone said, ignorance of history leads to repetition. The pro quoted in Bogiesan's post reflects my feelings about playing. The AI bots have introduced many new ideas and will change how the game is played, but actually playing involves two people, the one playing Black and the one playing White. The bots can function like a fitness center, playing against an AI is like lifting weights to develop your body's strength. Granted, some people like to do that for itself, but if you go that way you can end up muscle bound, with a body that can lift weights but not do much else.

It is well known that the best way to understand something is to figure it out for yourself. I think that applies to learning how to play go, as we are all always doing. The pro game records give us pointers as to what might be a good way to play and it is up to us to figure out from the pointers why that way might be good. If you can do that you may even decide that the pro move wasn't the best. AI bots also give us pointers and it is up to us to find out how to understand. In the end we have to play go and find out that way what we have learned from the various pointers we've gotten.

If you want to dispose of your books why not consider giving them to a go center? If I were experiencing a space problem and had to get rid of some books I'd do a triage. I probably would get rid of books that spoke to learners at a level much lower than my own. I would jettison problem books that I have studied so much that I know the solutions to all the problems at a glance. I would definitely keep the collections of pro games such as the 1971 Honinbo Tournament, and the Japanese books of selected games by famous players (Nihon Igo Taikei, and Gendai Igo Taikei. I always see new things going over pro games. Furthermore as the pro quoted above said, I also want to see go played from the heart.

Top
 Profile  
 
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #5 Posted: Fri Aug 23, 2019 7:25 am 
Gosei

Posts: 1733
Location: Earth
Liked others: 621
Was liked: 310
pro games database and AI are a natural combo in my unhumble opinion

Top
 Profile  
 
Offline
 Post subject: Re: Game collections, opening and Joseki books in the AI age
Post #6 Posted: Fri Aug 23, 2019 7:40 am 
Honinbo

Posts: 9545
Liked others: 1600
Was liked: 1711
KGS: Kirby
Tygem: 커비라고해
Before AI overcame humans, not that long ago, I recall that some folks advised against studying the games of Lee Sedol. The idea was that Lee Sedol played in a complicated way, and it was hard to really understand him, anyway. If you were going to study pro games, it'd be better to study pros that played in a simpler way that us amateurs could understand.

Lee Sedol's moves may have been good against top pros, but those moves were good in exceptional circumstances - they weren't the norm. Later, some of Lee's commentary books became popular, and this advise was widely disregarded. Also, some people just liked Lee Sedol, so they studied his games anyway. On the other hand, some folks preferred studying simpler games, because the moves they saw in those games may be more widely applicable to their own games.

Modern AI is comparable. Computers play at a level far beyond humans, and have an overall better quality of play. But for us amateurs, it may be difficult to understand the meaning behind AI moves that vary from what human pros of the past would have played.

Now let's step back and look at amateur games. On this L19 forum, folks post their own games for analysis and review. We look at them, draw observations, and think about what moves were good or bad. This, too, is educational.

My takeaway is that games from all levels - AI, pros, and amateurs - can be educational, though, perhaps in different ways. Value can be found in all of this, given that you're actively thinking about go.

So let's study it all, each according to one's individual taste. :salute:

_________________
be immersed

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 6 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group